Synthesis and discovery of a drug candidate for treatment of idiopathic pulmonary fibrosis through inhibition of TGF-β1 pathway

Article abstract of DOI:10.1016/j.ejmech.2018.07.074

In this study, anti-IPF lead compounds 42 and 44, derived from natural sesquiterpene lactones Isoalantolactone and alantolactone, were discovered by screening from a high-throughput TGF-β1 reporter luciferase assay. Notably, they could reduce the myofibroblast activation and extracellular matrix deposition both in vitro and in vivo. Additionally, compounds 42 and 44 could significantly attenuate bleomycin-induced pulmonary fibrosis in mice. Further validation of pharmacokinetics study and toxicity evaluation indicated that compound 44 might be a promising anti-IPF drug candidate.

Full text of DOI:10.1016/j.ejmech.2018.07.074

Accepted Manuscript
Synthesis and discovery of a drug candidate for treatment of idiopathic pulmonary
fibrosis through inhibition of TGF-β1 pathway
Xiaohe Li, Cheng Lu, Shuangwei Liu, shuaishuai Liu, Chengcheng Su, Ting Xiao,
Zhun Bi, Pengzhen Sheng, Mengying Huang, Xinhua Liu, Yujiao Wei, Lin Zhao,
Shengxiang Miao, Jiahe Mao, Kai Huang, Shaoyan Gao, Ning Liu, Min Qi, Tongtong
Liu, Shuanglin Qin, Luqing Wei, Tao Sun, Wen Ning, Guang Yang, Honggang Zhou,
Cheng Yang
PII:
S0223-5234(18)30639-1
10.1016/j.ejmech.2018.07.074
EJMECH 10609
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 7 June 2018
Revised Date: 30 July 2018
Accepted Date: 31 July 2018
Please cite this article as: X. Li, C. Lu, S. Liu, shuaishuai Liu, C. Su, T. Xiao, Z. Bi, P. Sheng, M. Huang,
X. Liu, Y. Wei, L. Zhao, S. Miao, J. Mao, K. Huang, S. Gao, N. Liu, M. Qi, T. Liu, S. Qin, L. Wei, T.
Sun, W. Ning, G. Yang, H. Zhou, C. Yang, Synthesis and discovery of a drug candidate for treatment
of idiopathic pulmonary fibrosis through inhibition of TGF-β1 pathway, European Journal of Medicinal
Chemistry (2018), doi: 10.1016/j.ejmech.2018.07.074.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Synthesis and Discovery of a Drug Candidate for Treatment of Idiopathic Pulmonary
Fibrosis through Inhibition of TGF-β1 Pathway

ACCEPTED MANUSCRIPT
Synthesis and Discovery of a Drug Candidate for Treatment of Idiopathic Pulmonary
Fibrosis through Inhibition of TGF-β1 Pathway
1
, §
1, §
1, §
1, §
2
Xiaohe Li, Cheng Lu, Shuangwei Liu, shuaishuai Liu, Chengcheng Su, Ting
1
1
1
1
1
1
Xiao, Zhun Bi, Pengzhen Sheng, Mengying Huang, Xinhua Liu, Yujiao Wei, Lin
1
1
1
1
1
1
1
Zhao, Shengxiang Miao, Jiahe Mao, Kai Huang, Shaoyan Gao, Ning Liu, Min Qi,
1
1
2
1
1
*,1
Tongtong Liu, Shuanglin Qin, Luqing Wei, Tao Sun, Wen Ning, Guang Yang,
*,1
*,1
Honggang Zhou, Cheng Yang
1
College of Pharmacy, The State Key Laboratory of Medicinal Chemical Biology, and
Tianjin Key Laboratory of Molecular Drug Research, Nankai University, Tianjin 300350,
People’s Republic of China
2
Department of Respiratory and Critical Care Medicine, Pingjin Hospital, Logistics
University of Chinese People’s Armed Police Forces, Tianjin, 300162, People’s
Republic of China
Keywords
Idiopathic Pulmonary Fibrosis
TGF-β1/p-Smad3 Signaling Pathway
Sesquiterpene Lactone
Isoalantolactone
High-throughput Screening
ABSTRACT
In this study, anti-IPF lead compounds 42 and 44, derived from natural sesquiterpene
lactones Isoalantolactone and alantolactone, were discovered by screening from a
high-throughput TGF-β1 reporter luciferase assay. Notably, they could reduce the
myofibroblast activation and extracellular matrix deposition both in vitro and in vivo.
Additionally, compounds 42 and 44 could significantly attenuate bleomycin-induced
pulmonary fibrosis in mice. Further validation of pharmacokinetics study and toxicity

ACCEPTED MANUSCRIPT
evaluation indicated that compound 44 might be a promising anti-IPF drug
candidate.
INTRODUCTION
Idiopathic pulmonary fibrosis (IPF) is a chronic, progressive, and often fatal lung
disease of unknown etiology. It is characterized by myofibroblasts accumulation and
[
1]
extracellular matrix (ECM) deposition. A variety of causes could lead to IPF, while
the mechanism of the underlying pathologic process remains poorly understood. IPF
is associated with high morbidity and mortality, with a mean survival ranging from
[
2]
3
.2 to 5 years after diagnosis. However, current therapies are largely ineffective
against the disease progression, and most clinical trials of new therapies designed for
[3]
IPF have obtained negative results in the past decades.
Myofibroblasts are the key effector cells in pulmonary fibrogenesis and the
primary cell type responsible for the synthesis and deposition of ECM including
[
4]
collagens I (COL1) and fibronectin (Fn).
Various cytokines and growth factors
within the tissue microenvironment contribute to the activation of myofibroblasts.
Among them, transforming growth factor (TGF)-β has been implicated as a “master
[
5, 6]
switch” in the induction of lung fibrosis and myofibroblasts differentiation.
The
role of TGF-β1 is well described in IPF. To be specific, TGF-β1 is increased in tissue
[
7]
[8]
samples from both bleomycin-induced animal model and IPF patients. Moreover,
over-expression of an adenovirus encoding active TGF-β1 leads to persistent
[
9]
[10]
pulmonary fibrosis, while inhibition of TGF-β1 with soluble TGF-β receptor,
or a
ameliorates pulmonary fibrosis. In
addition, mice null for the TGF-β1 downstream signaling p-Smad3 are protected from
[11]
TGF-β receptor 1 (TβRI, ALK5) inhibitor,
[
12]
pulmonary fibrosis.
All together, TGF-β1 plays a crucial role in the pathogenesis of
IPF. Based on this theory, we established a NIH3T3 fibroblasts cell line stably
transfected with (CAGA) -Lux-reporter, which encoded 12 copies of canonical Smad3
1
2
DNA-binding sequence CAGA. This CAGA-NIH3T3 screening system was utilized for
high-throughput screening compounds that can effectively inhibit the TGF-β signaling
pathway and further attempt to discover the leading compounds for treating
pulmonary fibrosis.

ACCEPTED MANUSCRIPT
Previous studies have indicated that natural sesquiterpene lactone compounds
[
13-15]
harbored extensive connection with TGF- β1 signaling pathway.
Therefore, it
might be a good resource to discover anti-IPF compounds from sesquiterpene
lactones and their analogues.
Alantolactone (1) and isoalatolactone (2), belonging to a family of naturally
occurring unique eudesmane-type sesquiterpene lactone, are enriched in the
traditional Chinese medicine Inula helenium L (Shown in Scheme 1). Both of these
two natural products exhibit considerable pharmacological activities, including
[16-20]
antitumor, antibacterial, and anti-inflammatory effects.
Furthermore,
isoalatolactone has been shown to attenuate the high glucose-stimulated activation
of TGF-β1 and expression of monocyte chemoattractant protein-1 (MCP-1) and
[
21]
fibronectin (Fn) in mesangial cells of rats.
However, the relationship between
isoalatolactone and TGF-β1 in the progression of IPF has not been investigated.
Herein, semi-synthetic analogues were prepared based on the structure of
alantolactone and isoalatolactone, followed by investigation of their inhibition on
TGF-β1 signaling pathway and the potency for treatment of IPF.
Scheme 1. Structure of Alantolactone (1) and Isoalatolactone (2)
RESULTS AND DISCUSSION
Chemistry
We took advantage of 1 and 2, naturally abundant sesquiterpene lactones, as
basic scaffold to synthesize the derivatives (Scheme 2). First of all, alantolactone (1)
was introduced an epoxide ring via simple oxidization by m-CPBA to obtain a stable
molecule 3. Secondarily, isoalantolactone (2) was converted into intermediates 4 by
mixed with SeO in TBHP. Subsequently, the resulting allylic alcohol 4 was oxidized by
2
m-CPBA to yield compound 6.

ACCEPTED MANUSCRIPT
H
H
H
O
O
a
O
O
O
H
1
3
H
H
H
H
H
O
b
O
c
O
O
O
O
HO
HO
H
H
H
O
H
2
4
6
d
H
O
O
O
H
H
5
Reagents and conditions: a) m-CPBA, CH Cl , rt, 90%; b) SeO , TBHP, reflux, 65%; c)
2
2
2
m-CPBA, CH Cl , rt, 85%; d) m-CPBA, CH Cl , rt, 92%.
2
2
2
2
Scheme 2. Preparation of analogues 3, 4, 5 and 6
With the key intermediates 4 in hands, functional groups transformation could
provide more synthetic analogues 7-12 for biological investigation (Scheme 3).
Specifically, pretreatment of compound 4 with NaH, following by reaction with BnBr
to give rise to analogue 7 (Table 1, Entry 1). The secondary alcohol 4 was mixed with
TBSCl and TBDPSCl, in the presence of imidazole, delivered the silyl ethers 8 and 9
(Table 1, Entries 2 and 3). This alcohol 4 was oxidized by Dess-Martin periodinane to
furnish the enone 10.
H
H
H
H
O
O
a or b or c
O
O
HO
H
1
R O
H
O
4
7-9
H
H
H
H
H
H
O
e
O
d
O
O
O
HO
H
O
N
H
H
OH
4
10
11
f
H
O
O
H N
2
H
H
12
o
Reagents and conditions: a) NaH, DMF, BnBr, 0 C to rt, 78%; b) TBSCl, imidazole, DMF,
o
rt, 85%; c) TBDPSCl, imidazole, DMF, rt, 85%; d) Dess-Martin periodinane, CH Cl , 0 C
2
2
.
o
to rt, 65%; e) Py., NH OH HCl, MeOH, reflux, 72%; f) NH Cl, MeOH, NaCNBH , 0 C to
2
4
3

ACCEPTED MANUSCRIPT
rt, 43%.
Scheme 3. Synthetic Approach to analogues 7-12.
Table 1. Structures of Analogues 7-9
1
Entry
No. of Compounds
R
1
2
3
7
8
9
Bn
TBS
TBDPS
Also, we tried to prepare various esters analogues 13-40 on the secondary
alcohols of the intermediates 4 and 6 (Table 2). Under our optimized conditions
(Scheme 4), esterification could process efficiently in the presence of DIC and
catalytic amount of DMAP in polar solvents (Such as MeCN or DMF).
Reagents and conditions: a) RCO H, DIC, DMAP, DMF or MeCN, rt, 28%-95% yield.
2
Scheme 4. Preparation of the Ester Analogues 13-41
Table 2. Structures of Analogues 7-9
2
2
Entry No.
of R
Yield/% Entry No.
of R
Yield/%
40
Compounds
Compounds
1
2
13
55
96
16
17
28
14
29
85
3
4
15
16
84
93
18
19
30
31
63
55

ACCEPTED MANUSCRIPT
5
6
7
8
9
1
1
17
18
19
20
21
22
23
92
77
47
56
68
75
84
20
21
22
23
24
25
26
32
33
34
35
36
37
38
28
71
56
91
58
46
39
0
1
1
1
1
2
3
4
24
25
26
38
29
35
27
28
29
39
40
41
44
42
51
1
5
27
39
Biology
Screening of functional compounds targeting TGF-β1/Smad3 signaling pathway
in fibroblasts. The inhibitory potency of Alantolactone and Isoalatolactone analogues
on TGF-β1/Smad3 signaling pathway in fibroblasts was evaluated by CAGA-luciferase
report gene assay as described in experimental section. Fold changes of luciferase
were shown in Figure 1, and the inhibition ratios of compounds were summarized in
Table 3. Firstly, the natural products alantolactone 1 and isoalantolactone 2 were not
found to exert any inhibition effect on TGF-β1/Smad3 signaling. Compared to natural

ACCEPTED MANUSCRIPT
alantolactone 1, the exposide analogue 3 exhibited a remarkable growth inhibition
on TGF-β1/Smad3 pathway, with an inhibitory rate of 77%, which was the most
potent analogues. Hydroxylated analogue 4 and 6 of isoalantolactone also exerted a
stronge effect on blocking the TGF-β1/Smad3 pathway (61% and 46% of inhibitory
rate, respectively). However, the analogues 5, 10, 11 and 12 of isoalantolactone did
not shown significant inhibitory effect against the signaling pathway activated by
TGF-β1, indicating that the hydroxyl group at C-3 might be the essential group to
maintain the activity on this signaling pathway. Further modification was focused on
the fuctionalization of this hydroxyl group. An ether and silyl ether analogues 7, 8
and 9 reduced the inhibitory acvitivity. And the inhibition activity of most ester
analogues was abolished except compound 38 and 41. These two ester analogues
harbored potent inhibitory rate on Smad3-activation (71% and 34% of inhibitory rate,
respectively).
Figure 1. Screening of functional compounds targeting TGF-β1/Smad3 signaling
pathway in fibroblasts.
CAGA-NIH3T3 cells were treated with/without TGF-β1 (5 ng/mL) accompanied with
compounds (10μΜ) for 18h. Fold change of luciferase were divided
TGF-β1-stimulated luciferase by luciferase in TGF-β1 non-stimulated group.
Experiments were performed in triplicate, and the statistical significance was
obtained with one-way ANOVA. Error bars indicate mean ± SEM. *P<0.05, * *P<0.01,
*
**P<0.001.
Table 3. Inhibition ratio of compounds on TGF-β1/Smad3 signaling pathway in
fibroblasts
compound Inhibition
compound Inhibition
compound Inhibition

ACCEPTED MANUSCRIPT
Ratio
Ratio
Ratio
1
2
3
4
5
6
7
8
9
1
1
1
1
1
-27.73%
-35.36%
77.72%
61.85%
3.08%
15
16
17
18
19
20
21
22
23
24
25
26
27
28
-23.10%
-49.83%
6.58%
29
30
31
32
33
34
35
36
37
38
39
40
41
-7.65%
6.65%
-190.46%
-16.96%
26.69%
-2.36%
-22.34%
17.11%
3.67%
4.63%
-32.69%
-202.26%
-74.69%
-187.25%
21.97%
-37.58%
-17.25%
-13.63%
-30.63%
-30.71%
46.00%
-29.47%
27.79%
-74.69%
2.24%
0
1
2
3
4
71.29%
13.46%
-136.64%
34.79%
-11.95%
-21.86%
-55.44%
-34.58%
CAGA-NIH3T3 cells were treated with/without TGF-β1 (5 ng/mL) and compounds (10
μΜ) for 18h. The inhibition ratio was calculated by the TGF-β1-stimulated luciferase
value of compounds-treated group compared with control group. All values are the
mean of three independent experiments.
Selected Compounds Inhibited TGF-β1-induced Proliferation of Lung
Fibroblasts. According to the results of luciferase screening, five compounds (3, 4, 6,

ACCEPTED MANUSCRIPT
3
8 and 41) were selected for further investgation of anti-proliferative effects on lung
fibroblasts. MTT assays indicated that compounds 3, 4, 6 and 38 suppressed
TGF-β1-stimulated proliferation of fibroblasts (Figure 2).
Figure 2. Selected compounds inhibited TGF-β1-induced proliferation of lung
fibroblasts.
Mlg cells were treated with/without TGF-β1 (5 ng/mL) and compounds (10 μΜ)
for 24h. The cells treated with DMSO were served as control. The five compounds,
selected in the luciferase reporter screening system, were measured by MTT assays
at 24h to determine the inhibitory effect of compounds on fibroblast proliferation,
which were activated by TGF-β1. All values were normalized to the control,
representing 100% cell viability. Experiments were performed in triplicate, and the
statistical significance was obtained with one-way ANOVA. Error bars indicate mean ±
SEM. *P<0.05.
Selected compounds inhibited TGF-β1-induced migration of lung fibroblasts. In
addition, motility potential of lung fibroblasts under TGF-β1 and compounds
treatment was further evaluated. Wound healing assays indicated that compounds 3,
4
, 6 and 38 could inhibit TGF-β1-induced migration of lung fibroblasts, while the
compound 41 exhibited weaker inhibition ability in comparison to other compounds
Figure 3).
(

ACCEPTED MANUSCRIPT
Figure 3. Selected compounds inhibited TGF-β1-induced migration of lung
fibroblasts.
Mlg cells were treated with/without TGF-β1 (5 ng/mL) and compounds (10 μΜ) for
1
2h or 24 h. (A) The images were photographed at 0h, 12h and 24h post scratching
(40x). (B) The wound areas were measured at the three time points in every group,
and distance ratio was compared with initial scratch area. Error bars indicate mean ±
SEM.
Selected compounds inhibited TGF-β1-induced activation of lung fibroblasts.
Progressive pulmonary fibrosis was characterized by the accumulation of activated
[
22]
fibroblasts (also named myofibroblasts) in the fibrotic foci. Thus, TGF-β1 was used
to activate lung fibroblasts, followed by investigation of whether the selected
compounds could interfere with the activation of fibroblasts. Real-time PCR assays

ACCEPTED MANUSCRIPT
were used to detect the mRNA levels of α-SMA, type I collagen (Col1a1) and
Fibronectin (Fn), the makers of newly appearing myofibroblasts after TGF-β1
stimulation (Shown in Figure 4). As shown in Figure 4, compounds 3 and 6 could
inhibit TGF-β1–induced mRNA expression of α-SMA; compounds 3, 4, 6 and 38 could
decresed mRNA levels of Col1a1; compounds 3, 4 and 6 could significantly suppress
mRNA expression levels of Fn in lung fibroblasts upon 24 h co-incubation. Together,
these data suggested that compounds 3, 4, 6, 38 and 41 were able to inhibit the
activation of lung fibroblasts.

ACCEPTED MANUSCRIPT
Figure 4. Selected compounds inhibited TGF-β1-induced activation of lung fibroblasts.
Lung fibroblasts were treated with/without TGF-β1 (5 ng/mL) and compounds (10
μΜ) for 24 h.
(A-C) Real-time PCR was performed to examine the mRNA level of α-SMA,
Col1a1 and Fn. Data are expressed as mean ±SEM, n=3, *P< 0.05, **P< 0.01.
Preparation of Water-soluble Prodrug Molecules 42-46

ACCEPTED MANUSCRIPT
Due to the poor PK data of compound 3, 4, 6, 38 and 41, water-soluble prodrugs
4
2-46 were synthesized based on previous strategy, which could improve the PK
[
23-25]
properties, dramaticly.
As shown in scheme 5, compounds 3, 4, 6, 38 and 41
reacted with dimethylamine via michael addition, followed by mixture with maleic
acid to gain stable salts 42-46 (shown in Scheme 5).
Reagents and conditions: a) Me NH.HCl, K CO , CH Cl , reflux; b) Fumaric acid, EtOAc,
2
2
3
2
2
rt, 58-89% over 2 steps.
Scheme 5. Preparation of the Water-soluble Analogues 42-46.
Water-soluble prodrugs 42-46 dose-dependently inhibited TGF-β1/Smad3
signaling pathway in fibroblasts. To investigate whether the water-soluble analogues
4
4
2-46 exerted an inhibitory effect on TGF-β1/smad3 signaling pathways, compounds
2-46 were gradient screened by a CAGA-luciferase reporter system.Consequently,
the luciferase assay illustrated that compounds 42 and 44 could significantly inhibit
TGF-β1/smad3 signaling in a dose-dependent manner (Figure 5A and 5C);
compounds 45 displayed high suppression ability at a concentration of 40 μΜ (Figure
5
D); compounds 43 and 46 showed moderate inhibition during gradiant screening
(Figure 5B and 5E). Collectively, these results suggested that water-soluble analogues

ACCEPTED MANUSCRIPT
4
2, 44 and 45 have relative higher biological potency compared with compound 43
and 46.
Figure 5. Inhibition activity detection of water-soluble analogues 42-46 on
TGF-β1/Smad3 signaling pathway in fibroblasts.
(A-E) CAGA-NIH3T3 fibroblasts cells were treated with/without TGF-β1 (5 ng/mL)
and different doses of water-soluble analogues 42-46 for 18h. Concentration
gradients of compounds were set as 5, 10, 20 and 40 μM. Experiments were
performed in triplicate, and the statistical significance was obtained with one-way
ANOVA. Error bars indicate mean ± SEM. *P<0.05, **P<0.01, ***P<0.001.

ACCEPTED MANUSCRIPT
Compounds 42, 44 and 45 inhibited fibroblasts activation and ECM production
in vitro through suppressing TGF-β1/Smad3 signaling pathway. According to the
outcomes of luciferase assay, compounds 42, 44 and 45 were chosen for further
biological activity verification. The inhibitory capability of compounds 42, 44 and 45
on TGF-β1 responsiveness was analyzed in lung fibroblasts. As expected, compounds
4
2, 44 and 45 could decrease TGF-β1-induced α-SMA expression and ECM production
in fibroblasts, accompanied by much lower phosphorylation levels of Smad3 (Figure
). These results suggested that the water-soluble analogues 42, 44 and 45 inhibited
fibroblasts activation possibly through suppressing TGF-β1/Smad3 signaling pathway.
6

ACCEPTED MANUSCRIPT
Figure 6. Compounds 42, 44 and 45 inhibited fibroblasts activation and ECM
production through suppressing TGF-β1/Smad3 signaling pathway.
(A) Lung fibroblasts were treated with/without TGF-β1 (5 ng/mL) and
compounds 42, 44 and 45 (10 μΜ) for 24 h. The expression of α-SMA in cell extracts,
Col1 in the medium (supernatant) were examined by Western blot. Lung fibroblasts
were treated with/without TGF-β1 (5 ng/mL) and compound 42, 44 and 45 (10 μΜ)
for 30min, the expression of p-Smad3 was examined by Western blot. β-Actin was
used as a loading control. (B-C) The relative density of the bands of α-SMA and
p-Smad3 normalized to β-actin is represented in the graphs. Data are presented as
mean ± SD (n = 4). *P < 0.05, ***P < 0.001.

ACCEPTED MANUSCRIPT
Compound 42, 44 and 45 attenuated bleomycin-induced pulmonary fibrosis in
mice. To further determine the theraputic impact of the compounds 42, 44 and 45
on lung fibrosis in vivo, bleomycin (BLM)-induced pulmonary fibrosis mice model was
established, which were treated with compounds 42, 44 and 45 after bleomycin
injury. Pirfenidone, a listed drug for the treatment of IPF, was used as a positive
control. The pathologic changes in the lung was shown in Figure 7A. The sections of
the NaCl group displayed clear lung structures while the alveolar structures were
destroyed in the model group. Compound 42, 44 and 45 treatments improved the
alveolar structure compared with that in the model group and Pirfenidone group.
The percentage of fibrotic area was reduced by approximately 71.8%, 60.5% and 53.5%
following the treatment of compounds 42, 44 and 45, respectively (Figure 7B).
Collagen content was decreased in the mice treated with compounds 42, 44 and 45,
compared with those treated with Pirfenidone. Among three compounds,
compounds 42 and 44 harbored a preferable effect on decreasing ECM production
(Figure 7C). In summary, the histopathological analysis and collagen content
suggested that compounds 42, 44 and 45 exerted a promising theraputic impact on
BLM-induced pulmonary fibrosis in mice.

ACCEPTED MANUSCRIPT
Figure 7. In vivo effects of compounds 42, 44 and 45 on bleomycin-induced
pulmonary fibrosis in mice.
C57BL/6 male mice were intratracheal instillated with BLM (2 U/kg) to induce
pulmonary fibrosis. Compounds 42, 44, 45 and Pirfenidone (all 100 mg/kg) were
given orally once daily from day 7-14 after BLM treatment and lungs were harvested
at day 14 for the following analysis (n = 10 per group). (A) Body weight of each group.
(B) Hydroxyproline contents in lung tissues of each group were measured as decribed
in the methods (*P < 0.05). (C) Lung fibrotic score analysis of the lung sections. The
fibrotic area is presented as a percentage. The lung fibrosis areas of each group were
calculated using Image-Pro Plus Version 6.0 software (Media Cybernetic, Inc.
American). ** P < 0.01. (D) 14 days after BLM instillation, lung histology of each
group mice was determined by hematoxylin-eosin (H&E) (200X).
Compounds 42, 44 and 45 inhibited ECM production in vivo through
suppressing TGF-β1/Smad3 signaling pathway. To explore the mechanism of

ACCEPTED MANUSCRIPT
compounds 42, 44 and 45 on BLM-induced fibrogenic changes, the protein levels of
collagen1 and p-Smad3 in lung tissues were assesses in each group. Consistent with
the hydroxyl-proline assay, the expression levels of col1 were significantly reduced in
compounds-treatment group (Figure 8A and 8B). The phosphorylation levels of
Smad3 in lung tissues of compounds 42 and 44 treated group were apparently
declined compared with those in other groups (Figure 8A and 8C). Consistently, the
immunohistochemical analysis of α-SMA, Col1 and p-Smad3 in lung sections
indicated the same results (Figure 8D and 8E). In consideration of the luciferase
results of compound 45, the lower anti-fibrosis potency of compound 45 may be
associated with the moderate inhibitory ability on TGF-β1/Smad3 signaling pathway
and a higher administration dose may improve the anti-fibrosis efficacy.

ACCEPTED MANUSCRIPT
Figure 8. Compounds 42, 44 and 45 inhibited ECM production in vivo through
suppressing TGF-β1/Smad3 signaling pathway.
C57BL/6 male mice were intratracheal instillated with BLM (2 U/kg) to induce
pulmonary fibrosis. Compounds 42, 44, 45 and Pirfenidone (all 100 mg/kg) were
given orally once daily from day 7-14 after BLM treatment and lungs were harvested
at day 14 for the following analyses (n = 10 per group). (A-C) The expression of
p-Smad3 and Col1 in lung tissues were determined by Western blot analysis, GAPDH
was used as a loading control. Relative density of the bands of p-Smad3 and Col1
normalized to GAPDH were represented in the graphs. (D-E) Immunohistochemical
analysis of α-SMA, Col1 and p-Smad3 in lung sections. Representative images of the
staining are shown (n = 10 per group). Error bars indicate mean ± SEM.

ACCEPTED MANUSCRIPT
Compounds 42 and 44 inhibited human lung fibroblasts activation and ECM
production through suppressing TGF-β1/Smad3 signaling pathway.
According to the results of in vivo experiments, compounds 42 and 44 showed a
significant anti-fibrosis potency in bleomycin-induced mice model. Afterwards, we
evaluated the impact of compounds 42 and 44 on human lung fibroblasts. As shown
in Figure 9, the administration with compounds 42 and 44 resulting in a blunted
TGF-β1-induced phosphorylation of Smad3, protein expression of α-SMA and
production of type I collagen in human lung fibroblasts. The above-described results
were consistent with both in vitro and in vivo experiments on mice, suggesting that
compounds 42 and 44 may be useful in clinical experiments.

ACCEPTED MANUSCRIPT
Figure 9. Compounds 42 and 44 inhibited human lung fibroblasts activation and ECM
production through suppressing TGF-β1/Smad3 signaling pathway.
(A) Human lung fibroblasts (HFL1) were treated with/without TGF-β1 (5 ng/mL)
and compounds 42 and 44 (10 μΜ) for 24 h. The expression of α-SMA in cell extracts,
Col1 in the medium (supernatant) were examined by Western blot. Fibroblasts were
treated with/without TGF-β1 (5 ng/mL) and compounds 42 and 44 (10 μΜ) for 30min,
the expression of p-Smad3 was examined by Western blot. β-Actin was used as a
loading control. (B-C) The relative density of the bands of α-SMA and p-Smad3
normalized to β-actin is represented in the graphs. Data are presented as mean ± SD
(n = 3). **P < 0.01, ***P < 0.001.
Stability in Mouse Plasma. The plasma stability of compounds 42 and 44 in rat
plasma for 12 h was analyzed (Shown in Figure 10). After incubation with plasma at
7℃ for 3 h, the remaining ratio of compound 42 and 44 was more than 50%
3
compared to the initial concentration, indicating that both compounds were stable in
rat plasma. Besides, within 12 h, compounds 42 and 44 continuously transformed
into 3 and 6, respectively, through a retro Michael reaction (Scheme 6). These results
showed that compounds 42 and 44 had superior kinetic properties as prodrugs.
Figure 10. Concentration−ꢀme curve of compounds 42 (A) and 44 (B) in rat plasma.

ACCEPTED MANUSCRIPT
Scheme 6. Compounds 42 and 44 Transformed into 3 and 6 through a Retro Michael
Reaction in Rat Plasma
Pharmacokinetic (PK) Study of compounds 42 and 44 in Mice. Pharmacokinetic
(
PK) Study of compounds 42 and 44 in Mice. The plasma concentration−ꢀme curves
for compounds 42 and 44 after a single dose in rats were shown in Figure 11. The
pharmacokinetic parameters determined with a non-compartmental model were
shown in supplementary material. The plasma concentrations of both compounds 42
and 44 rapidly reached peak, and gradually declined after PO administration. Besides,
compounds 42 and 44 were constantly releasing 3 and 6 in vivo, resulting in an
effective concentration range of the blood concentration of 3 and 6 for hours. The
pharmacokinetics of both compounds were investigated by tail intravenous injection
(IV, Figure 10A and figure 10B), and the pharmacokinetic parameters were shown in
supplementary material. The oral bioavailability of compound 42 was determined to
be 75%, while compound 44 had a better bioavailability of 78%. The pharmacokinetic
properties of compound 42 and 44 made them drug-like molecules for further
development.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Figure 11. PK study of compounds 42 and 44 in rats. (A) compound 42 plasma
concentration−ꢀme curve in SD rats aꢁer a single IV administraꢀon of drug (30
mg/kg, mean ± SD, n = 5); (B) compounds 42 plasma concentration−ꢀme curve in SD
rats after a single PO administration of drug (150 mg/kg, mean ± SD, n = 5); (C)
compounds 44 plasma concentration−ꢀme curve in SD rats aꢁer a single IV
administration of drug (30 mg/kg, mean ± SD, n = 5); (D) compound 42 plasma
concentration−ꢀme curve in SD rats aꢁer a single PO administraꢀon of drug (150
mg/kg, mean ± SD, n = 5);
Toxicity Evaluation of Compounds 42 and 44. As a preliminary toxicology study,

ACCEPTED MANUSCRIPT
C57/BL6 mice were treated with compounds 42 and 44. Firstly, animals were
administered with 42 or 44 by gavage in a single dose (n = 10). For both compound
4
2 and 44, no animals died at 1, 2, 3, 4 and 5 g/kg. Then a 7-day oral chronic toxicity
study was conducted at 2 g/kg with both compounds 42 and 44 (dosed one daily, n =
1
3
0). All animals survived in both compounds 42 and 44 groups. Throughout another
0-day dosing period (dosed one daily, 2 g/kg, PO, n = 10), no animal displayed any
systemic symptoms during the observation period in the compound 42 group, while
two animals died in the compound 42 group. Together, these data indicated that
compound 42 had superior safety in vivo by oral administration.
CONCLUSION
In this report, 44 semi-synthetic analogues of natural Sesquiterpene Lactones
were prepared, whose anti-IPF potency was evaluated with the high-throughput
TGF-β1 reporter luciferase assay in vitro. The lead compounds 42 and 44 exerted high
inhibitory activities on TGF-β1 signaling pathway and inhibited TGF-β1-induced
α-SMA expression and ECM production in vitro. Additionally, they exhibited
promising efficacy in a bleomycin-induced pulmonary fibrosis mice model.
Furthermore, the compounds 42 and 44 could significantly inhibit human lung
fibroblasts activation in vitro. Further PK study and toxicity evaluation illustrated that
lead compound 44 might be a promising anti-IPF drug candidate, which deserved
further investigation.
EXPERIMENTAL SECTION
The Synthesis of the compound 3-46.
Chemistry.
General Methods. Before their use, the solvents were distilled and dried using
standard methods (Purification of laboratory chemicals (Six edition), Wilfred L. F.
1
Armarego and Christina L. L. Chai). H NMR spectra were obtained by using a Bruker
AV 400. Chemical shifts are reported in parts per million (ppm) relative to either a

ACCEPTED MANUSCRIPT
13
tetramethylsilane internal standard or solvent signals. C NMR spectra were
recorded using a Bruker AV 400 spectrometer (100 MHz). Chemical shifts (δ) are
reported in parts per million measured relative to the solvent peak. High-resolution
mass spectra (HRMS) were obtained with a FTICR-MS (Ionspec 7.0T) spectrometer.
The purity of all of the tested compounds was greater than 95%, which was identified
by HPLC.
(1aR,2S,5aR,6aR,9aR,9bS)-2,5a-Dimethyl-9-methyleneoctahydro-2H-oxireno[2',3':4,4
a]naphtho[2,3-b]furan-8(9H)-one (3). Alantolactone (2.0 g, 8.6 mmol) was dissolved
in CH Cl (10 mL). m-CPBA (2.1 g, 10 mmol) in dichloromethane (10 mL) was added
2
2
dropwise to this solution. The mixture was stirred at room temperature for 3 h. Then
the reaction mixture was diluted with 30 mL saturated Na S O solution, washed
2
2 3
with saturated NaHCO solution, dried over anhydrous Na SO , filtered, and
3
2
4
evaporated to afford a crude product, which was purified by column chromatography
[
petroleum ether−EtOAc (5:2)] to obtain compound 3 as a white solid (1.9 g, 90%).
1
mp 164-166℃. H NMR (400 MHz, CDCl ) δ (ppm) 6.40 (s, 1H), 5.77 (s, 1H), 4.78 –
3
4
1
.55 (m, 1H), 3.66 (dt, J = 8.9, 2.8 Hz, 1H), 2.89 (d, J = 2.5 Hz, 1H), 1.92 – 1.64 (m, 3H),
1
3
.59 – 1.28 (m, 6H), 1.11 (s, 3H), 1.04 (dd, J = 7.8, 2.1 Hz, 3H). C NMR (100 MHz,
CDCl ) δ (ppm) 169.5, 136.6, 123.6, 75.0, 67.4, 61.1, 39.4, 37.6, 37.3, 37.0, 32.5, 29.4,
3
+
+
2
2
3.8, 18.0, 16.4. HR-MS (ESI) m/z: calcd for C H O Na [M + Na] 271.1305, found
15 20 3
71.1306.
(3aR,4aR,6R,8aR,9aR)-6-Hydroxy-8a-methyl-3,5-dimethylenedecahydronaphtho[2,3-
b]furan-2(3H)-one (4). To a solution of SeO (87.0 mg, 0.74 mmol) in CH Cl (5 mL)
2
2
2
was added TBHP (0.37 mL), and the mixture was stirred at 0 ℃ for 30 min.
Isoalantolactone (500 mg, 2.15 mmol) in CH Cl (5 mL) was then added dropwise to
2
2
the reaction mixture, and the mixture was stirred at room temperature for 24 h.
Then the reaction mixture was diluted with 8 mL saturated Na S O solution, and
2
2 3
extracted with CH Cl . The combined organic layers were washed with saturated
2
2
NaHCO solution, dried over anhydrous Na SO , and evaporated under vaccum. The
3
2
4

ACCEPTED MANUSCRIPT
crude product was purified by column chromatography [petroleum ether−EtOAc
1
(
3:1)] to obtain compound 4 (white solid, 347 mg, 65%). mp 144-145℃. H NMR (400
MHz, DMSO-d ) δ (ppm) 5.96 (s, 1H), 5.75 (s, 1H), 4.88 (d, J = 1.6 Hz, 1H), 4.72 (s, 1H),
6
4
.53 (td, J = 4.8, 1.6 Hz, 1H), 4.45 (t, J = 1.7 Hz, 1H), 4.12 (d, J = 2.9 Hz, 1H), 3.13 (ddd,
J = 12.0, 6.9, 5.1 Hz, 1H), 2.45 – 2.34 (m, 1H), 2.01 (dd, J = 15.4, 1.8 Hz, 1H), 1.73 –
1
3
1
1
1
.52 (m, 5H), 1.27 – 1.04 (m, 2H), 0.69 (s, 3H). C NMR (100 MHz, DMSO-d ) δ (ppm)
6
70.5, 152.0, 142.8, 120.6, 108.1, 77.0, 71.8, 40.9, 40.0, 39.8, 35.8, 34.2, 29.9, 27.1,
+
+
7.4. HR-MS (ESI) m/z: calcd for C H O Na [M + Na] 271.1305, found 271.1304.
1
5 20 3
(3aR,4aR,5R,8aR,9aR)-8a-Methyl-3-methylenedecahydro-2H-spiro[naphtho[2,3-b]fur
an-5,2'-oxiran]-2-one (5). Isoalantolactone (1.00 g, 4.30 mmol) was dissolved in
CH Cl (16 mL). m-CPBA (890 mg, 5.16 mmol) was added to this solution in batches.
2
2
The mixture was stirred at room temperature for 3 h. Then the reaction mixture was
diluted with 16 mL saturated Na S O solution, washed with saturated NaHCO
3
2
2
3
solution, dried over anhydrous Na SO , filtered, and evaporated to afford a crude
2
4
product, which was purified by column chromatography [petroleum ether−EtOAc (20:
1
1
)] to obtain compound 5 (white solid, 984 mg, 92%). mp 124-126℃. H NMR (400
MHz, CDCl ) δ (ppm) 6.03 (s, 1H), 5.48 (d, J = 2.2 Hz, 1H), 4.41 (t, J = 5.0 Hz, 1H), 2.83
3
(dt, J = 11.6, 5.7 Hz, 1H), 2.60 (d, J = 4.2 Hz, 1H), 2.48 (t, J = 3.1 Hz, 1H), 2.12 (d, J =
1
3
1
5.6 Hz, 1H), 1.83 -1.09 (m, 9H), 0.95 – 0.70 (m, 4H). C NMR (100 MHz, CDCl ) δ
3
(ppm) 170.4, 141.8, 120.4, 76.5, 58.5, 50.7, 44.2, 41.9, 41.4, 40.4, 35.3, 34.3, 23.1,
+
+
2
2
0.3, 18.6. HR-MS (ESI) m/z: calcd for C H O Na [M + Na] 271.1305, found
15 20 3
71.1307.
(3aR,4aR,5R,6R,8aR,9aR)-6-Hydroxy-8a-methyl-3-methylenedecahydro-2H-spiro[nap
htho[2,3-b]furan-5,2'-oxiran]-2-one (6). Compound 4 (112 mg, 0.45 mmol) was
dissolved in CH Cl (1 mL). m-CPBA (100 mg, 0.60 mmol) in CH Cl (1 mL) was added
2
2
2
2
to this solution dropwise. The mixture was stirred at room temperature for 2 h. Then
the reaction mixture was diluted with 3 mL saturated Na S O solution, washed with
2
2 3
saturated NaHCO solution, dried over anhydrous Na SO , filtered, and evaporated to
3
2
4
afford a crude product, which was purified by column chromatography [petroleum