Design, synthesis and 3D-QSAR analysis of novel thiopyranopyrimidine derivatives as potential antitumor agents inhibiting A549 and Hela cancer cells

Article abstract of DOI:10.1016/j.ejmech.2019.111809

Four series of thiopyranopyrimidine AZD9291 derivatives containing acrylamide structure were designed, synthesized and evaluated for their antiproliferative activity against A549 and Hela cancer cells. Most of the compounds exhibited excellent antiproliferative activity against A549 cells. Moreover, the compounds with indole ring fluorine substituted exhibited better antiproliferative activity against Hela cells. The most promising compound 23g exhibited excellent enzymatic inhibitory activity and selectivity for EGFR^L858R/T790M double mutations. The IC₅₀ value against EGFR^L858R/T790M kinase was 16 nM. The compound 23g inhibits selectively against the mutated form of EGFR, with the selectivity more than 125-fold. Furthermore, compound 23g also inhibited A549 cells, Hela cells and H1975 cells proliferation at a low concentration, and the IC₅₀ values were 0.057 μM, 0.104 μM and 0.916 μM, respectively. To further investigate the QSARs of thiopyranopyrimidine derivatives, the CoMFA (q [2] = 0.765, r² = 0.965) and CoMSIA (q [2] = 0.875, r² = 0.956) models on Hela cancer cells were established. The generated 3D-QSAR model was validated to be reliable and can be used for further design and optimization of novel and selective EGFR inhibitors.

Full text of DOI:10.1016/j.ejmech.2019.111809

Journal Pre-proof
Design, synthesis and 3D-QSAR analysis of novel thiopyranopyrimidine derivatives
as potential antitumor agents inhibiting A549 and Hela cancer cells
Bingbing Zhao, Chengwu Zhao, Xiaohan Hu, Shan Xu, Zhou Lan, Yuping Guo,
Zunhua Yang, Wufu Zhu, Pengwu Zheng
PII:
S0223-5234(19)30961-4
DOI:
https://doi.org/10.1016/j.ejmech.2019.111809
EJMECH 111809
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 16 July 2019
Revised Date: 17 September 2019
Accepted Date: 21 October 2019
Please cite this article as: B. Zhao, C. Zhao, X. Hu, S. Xu, Z. Lan, Y. Guo, Z. Yang, W. Zhu, P. Zheng,
Design, synthesis and 3D-QSAR analysis of novel thiopyranopyrimidine derivatives as potential
antitumor agents inhibiting A549 and Hela cancer cells, European Journal of Medicinal Chemistry
(2019), doi: https://doi.org/10.1016/j.ejmech.2019.111809.
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.
© 2019 Published by Elsevier Masson SAS.

Design, synthesis and 3D-QSAR analysis of novel thiopyranopyrimidine derivatives as potential antitumor agents
inhibiting A549 and Hela cancer cells
Bingbing Zhaoa,b,#, Chengwu Zhaoc,#, Xiaohan Hua, Shan Xua,*, Zhou Lana, Yuping Guoa, Zunhua Yangb, Wufu Zhua,*,
Pengwu Zhenga,*
aJiangxi Provincial Key Laboratory of Drug Design and Evaluation, School of Pharmacy, Jiangxi Science & Technology
Normal University, 605 Fenglin Road, Nanchang, Jiangxi 330013, China
bCollege of Pharmacy, Jiangxi University of Traditional Chinese Medicine, Nanchang 330004, China,
cDepartment of Sports Medicine, The First Hospital of Jilin University, 71 Xinmin Road, Changchun, Jilin 130021, China
Abstract：
Four series of thiopyranopyrimidine AZD9291 derivatives containing acrylamide structure were designed,
synthesized and evaluated for their antiproliferative activity against A549 and Hela cancer cells. Most of the compounds
exhibited excellent antiproliferative activity against A549 cells. Moreover, the compounds with indole ring fluorine
substituted exhibited better antiproliferative activity against Hela cells. The most promising compound 23g exhibited
excellent enzymatic inhibitory activity and selectivity for EGFRL858R/T790M double mutations. The IC50 value against
EGFRL858R/T790M kinase was 16 nM. The compound 23g inhibits selectively against the mutated form of EGFR, with the
selectivity more than 125-fold. Furthermore, compound 23g also inhibited A549 cells, Hela cells and H1975 cells
proliferation at a low concentration, and the IC50 values were 0.057 µM, 0.104 µM and 0.916 µM, respectively. To further
investigate the QSARs of thiopyranopyrimidine derivatives, the CoMFA (q2 = 0.765, r2 = 0.965) and CoMSIA (q2 = 0.875,
r2 = 0.956) models on Hela cancer cells were established. The generated 3D-QSAR model was validated to be reliable
and can be used for further design and optimization of novel and selective EGFR inhibitors.
Keywords: Thiopyranopyrimidine; EGFR; inhibitor; 3D-QSAR; CoMFA; CoMSIA
1. Introduction
Epidermal growth factor receptor (EGFR) is widely expressed in most human cells and plays a vital role in maintaining
normal life activities1-3. Ras-Raf-MARK signaling pathway, PI3K/Akt/mTOR signaling pathway and JAK-STAT
*Corresponding authors.
E-mail: zhengpw@126.com (P. Zheng), zhuwufu-1122@163.com (W. Zhu), shanxu9891@126.com (S. Xu);
#These authors made equal contributions to this work.

signaling pathway are the main EGFR downstream signaling pathways reported in the litereatures4,5. These signaling
pathways coordinate with each other and jointly control a series of life activities such as cell survival, proliferation and
differentiation6,7. It was reported by the World Health Organization (WHO) that the incidence of overexpression or
mutation of EGFR signaling pathway in lung cancer, cervical cancer, breast cancer and other malignant tumor cells was
as high as 80%8. Consequently, deregulation or mutation of EGFR signaling pathway will lead to the occurrence of
diseases or malignancy.
Lung cancer ranks the first place in the world's total cancer mortality rate. The annual incidence of lung cancer is as
high as 1.8 million and the death toll is close to 1.6 million9,10. More than 85% of lung cancer patients belong to
non-small cell lung cancer (NSCLC) sub-types, with a 5-year survival rate of less than 10%11-13. The researches have
shown that the EGFR kinase mutations were the leading cause of NSCLC, for example, the T790M mutation14, L858R
mutation15, C797S mutation16 and other mutations17. In the past decades, a number of EGFR-target kinase inhibitors (1,
Gefitinb; 2, Afatinib and 3, AZD9291) have been approved by Food and Drug Administration (FDA) for treatment of
EGFR mutations in NSCLC patients18-20. As the first FDA-approved third-generation irreversible EGFR inhibitor,
AZD9291 has high selectivity for EGFRT790M mutation and remains the most widely used medicine for clinical treatment
of EGFRT790M mutation and advanced metastatic NSCLC21,22. However, the clinical therapeutic effect of AZD9291 is not
satisfactory. Some patients developed drug-resistant mutation (C797S) and serious side effects (cardiotoxicity, diarrhea
and rash) after treated with AZD929122.
F
Cl
F
Cl
N
O
N
O
HN
O
N
HN
N
HN
N
N
O
O
O
HN
O
N
N
N
N
N
N
O
H
1
2
3
O
N
O
O
N
N
N
O
OH
O
S
O
N
NH
N
S
N
S
N
N
O
N
N
N
H
N
6
O
N
5
N
N
4
O
S
H
H
BMC-20142422-7e
O
Figure 1 Representative EGFR inhibitors (1–3) and PI3k/mTOR inhibitors (4-6).
However, in addition to lung cancer, cervical cancer is also the cancer with the highest mortality in the world,
accounting for 33% of the total number of female cancer deaths, and more than 100,000 women die of cervical cancer
every year worldwide23. There is no effective medicine can cure cervical cancer completely. Human papillomavirus
(HPV) vaccine is the most effective vaccine for proventing cervical carcinoma at present24. But it still has many safety

problems (skin allergy and peripheral neuropathy). The disorder in PI3K/mTOR signal pathway was considered to be the
main cause of cervical cancer25. In recent years, many pyrimidine heterocyclic derivatives were reported as
PI3K-Akt-mTOR signal pathway inhibitors (4 BMC-20142422-7e, 5, 6)26-28. The compound 5 (GDC0941) and
compound 6 (PF-05139962) were effective PI3K signaling pathway inhibitors, which can significantly inhibit the growth
of cervical cancer cells (Hela). The IC50 values of compound 5 and 6 against PI3K kinase were 3 nM and 5 µM,
respectively27,28. The compound 4 (BMC-20142422-7e) had a good inhibitory ability on mTOR kinase, with the IC50
value of 0.27 µM26. These inhibitors have achieved remarkable efficiency in pre-clinical trials and are currently being
studied in the clinical trial. Consequently, the development of EGFR inhibitors based on EGFR signaling pathways will
greatly improve the accuracy of drug design and shorten the drug development of period.
In this research, the structure-based drug design (SBDD) strategy was applied to further study the structure-activity
relationship of lead compounds. Based on the structures of the reported potent leading compounds, aliphatic heterocycles
were introduced into pyrimidine ring by expansion principle and combination principle. Finally, four series of
thiopyranopyrimidine derivatives containing acrylamide structure (21a-g, 22a-g, 23a-g and 24a-g) were designed and
synthesized. It is desirable to obtain EGFR inhibitors with highly selectivity and potent inhibitory activities against
EGFRT790M mutant kinase. In addition, it is also expected to obtain inhibitors with dual cancer cells inhibitory activities of
A549 cancer cells and Hela cancer cells.
2. Results and discussion
2.1. Chemistry
Scheme 1 Synthetic routes of the compounds 13a-b and 14a-b. Reagents and conditions: (a) NaH, THF, rt., 2 h; (b) Urea，
C2H5ONa, C2H5OH, 70
ºC, 2.5 h; (c) POCl3, 100 ºC, 3 h; (d) CH3I, NaH, THF, rt., 1.5 h; (e) AlCl3, DME, 80 ºC, 3 h; (f)
Na2WO4†2H2O, H2O2, THF, rt., 0.5 h.
The synthetic routes of the compounds 13a-b and 14a-b were shown in Scheme 1. Commercially available starting
material 5-fluoroindole 7a or indole 7b reacted with iodomethane to get 8a-b. 10 was obtained by the reaction of

self-ester condensation from commercially available 9. Next, 10 cyclizated with urea to obtain 11. Then, 11 reacted with
phosphorus oxychloride to obtain intermediate 12. Intermediate 12 and 8a-b undergone Friedel-Crafts alkylation reaction
to obtain intermediate 13a-b. 14a-b were obtained by the reaction of oxidation from 13a-b.
Scheme 2 Synthetic routes of the compounds 21a-g, 22a-g, 23a-g and 24a-g. Reagents and conditions: (a) TsOH,
1,4-dioxane, 85
ºC
, 3 h; (b) N,N,N’-trimethylethylenediamine, DIPEA, DMA, 110 , 2.5 h; (c) Zn, NH4Cl, C2H5OH,
º
C
H2O, 105 , 3 h; (d) Oxalyl chloride, DMF, DCM, 0
º
C
ºC, 0.5 h; (e) NaHCO3, DCM, 0 ºC, 0.5-2 h.
The synthetic routes of the compounds 21a-g, 22a-g, 23a-g and 24a-g were shown in Scheme 2. The starting material
4-fluoro-2-methoxy-5-nitroaniline 15 reacted with intermediates 13a-b and 14a-b to get 16a-d. 18a-d were got by the
reaction of reduction from 17a-d with 16a-d as raw materials. Different substituted acid chloride (20a-g) were got from
the corresponding acrylic acids (19a-g) by the reaction of chlorination. Finally, 20a-g reacted with 18a-d to obtain the
target compounds 21a-g, 22a-g, 23a-g and 24a-g.
2.2. Docking study of lead compounds
The silico molecular docking strategy was applied to study the binding modes between compound skeletons and receptor
proteins. Sybyl 8.1 software (Tripos Associates Inc., USA) was used to simulate the binding patterns of lead compounds
AZD9291 and BMC-20142422-7e27 (4, Figure 1) with EGFRT790M protein (PDB code: 3IKA, www.rcsb.org). The
Surflex-Dock Geom was used to optimize the docking model results. Only the model with the best CS score was selected
to generate protein-ligand complexes for further molecular docking research. As depicted in Figure 2, the whole
AZD9291 skeleton embedded into hydrophobic pocket of EGFRT790M protein with a “U-shape” structure (Figure 2C) and
closely combined with EGFRT790M protein. The pyrimidine ring was far away from gatekeeper residue Met790. The
pyrimidine ring and the amino bridge formed a bidentate hydrogen bond with hinge region residue Met793.
The
acrylamide side chain extended to the conserved region residue Cys797 and formed a hydrogen bond with residue
Cys797. As illustrated in the Figure 2B, the compound BMC-20142422-7e also combined well with EGFRT790M protein

and formed a strong bidentate hydrogen bond with hinge region residue Met793. The phenolic hydroxyl structure
extended to the conserved residue Cys797, but without hydrogen bond formed. The orientation of the whole molecular
skeleton of BMC-20142422-7e in the EGFR protein cavity was basically the same as that of AZD9291(Figure 2D), and
the thiopyranopyrimidine ring also formed hydrogen bonding forces with EGFR protein. The molecular simulation
results further indicated that the thiopyranopyrimidine structure can be used to design novel EGFR inhibitors.
B
A
D
C
Figure 2 Binding site of leading compounds. (A) The docking result of AZD9291 (green) bound to EGFRT790M protein
(PDB code:3IKA). (B) The docking result of BMC-20142422-7e (cyan) bound to EGFRT790M protein. (C) The spatial
orientation of AZD9291 (green) bound to EGFRT790M protein. (D) The patial orientation of BMC-20142422-7e (cyan)
bound to EGFRT790M protein.
2.3 Structure-based drug design strategy
First of all, we conducted the computer simulation docking of the leading compounds. According to the docking results
(Figure 2), we can know that the pyrimidine ring of AZD9291 and the thiopyranopyrimidine ring of BMC-20142422-7e
all formed a bidentate hydrogen bond with the hinge region residue Met793 of EGFRT790M protein. It was clearly that any
modifications to the structures of those two parts will change the biological activity of the target compounds. The
thiopyranopyrimidine structure of PI3K/mTOR inhibitor BMC-20142422-7e formed a bidentate hydrogen bond with
EGFRT790M protein, which inspired us to firstly introduce aliphatic heterocycle into pyrimidine ring of AZD9291.
Through this strategy, the space volume of pyrimidine ring could be further expanded to fully occupy the protein cavity.
In addition, it is possible to retain the PI3K/mTOR kinase inhibitory activity of thiopyranopyrimidine skeleton. Secondly,
the halogen and small molecular alkyl side chains were introduced in the ortho position of acrylamide side chain of
AZD9291 to weaken the covalent binding ability of the target compounds to non-target proteins. In this way, the

cytotoxicity and side effects of the target compound will be greatly reduced. Finally, the electron-withdrawing group
F-atom was introduced into indole ring to change the electron cloud distribution of the indole ring and improve the
overall physical and chemical properties of the compound. Thus, it is hopeful to discover novel inhibitors with selectivity
and potent inhibition over EGFRT790M protein, and also expect to discover inhibitors with dual cancer cells inhibitory
activities.
Figure 3 Structure based drug design strategy from chemical structures of lead compounds.
2.4. Biological evaluation
2.4.1. In vitro antiproliferative activity against A549 and Hela cells
The in vitro antiproliferation activity of compounds 21a-g, 22a-g, 23a-g and 24a-g were shown in Table 1. As expected,
most of the thiopyranopyrimidine derivatives retained the excellently antiproliferative activity against A549 cells with the
IC50 values range from 0.057 µM to 7.568 µM. To our surprise, the majority of thiopyranopyrimidine derivatives also
exhibited potent antiproliferation activity against Hela cells, especially compounds with 5-fluoroindole structure (the IC50
values of compounds against Hela cells range from 0.175 µM to 7.096 µM). In general, compounds containing
5-fluoroindole structure (21a-g and 24a-g) had a higher antiproliferative activity against Hela cells than that of
compounds (22a-g and 23a-g) without fluorine-substituted indole ring structure. Furthermore, compounds (23a-g and
24a-g) containing a thiopyran dioxides ring structure (the LogD7.4 values range from 1.8 to 3.1) had a better LogD7.4
value than that of compounds (21a-g and 22a-g) without thiopyran ring double oxidated (the LogD7.4 values range from
3.8 to 5.2). However, the antiproliferative activity show that the double oxidation strategy had little effect on the
antiproliferative activity of the target compound. As the volume of acrylamide side chain increase, the antiproliferative
activity of the target compounds against A549 cells gradually increased. But the antiproliferative activity of the target
compounds against Hela cells remained unchange or even decrease slightly. The introduction of Cl-atom on the

acrylamide was beneficial to increase the antiproliferative activity of the target compound against A549 cells. The IC50
values of compounds 21a, 22a, 23a and 24a against A549 cells were 4.545 µM, 0.506 µM, 8.685 µM and 4.635 µM,
respectively. And the IC50 values of chlorine substituted compounds 21b, 22b, 23b and 24b against A549 cells were
0.864 µM, 0.716 µM, 00.881 µM and 3.183 µM, respectively. However, as shown in Table1, the introduction of Cl-atoms
was not conducive to increase the biological activity of the compounds against Hela cells and the antiproliferative
activities of compounds 21b,22b, 23b and 24b against Hela cells were weaker that of compounds 21a,22a, 23a and 24a.
The preferred compounds 23b (the IC50 values against A549 and Hela cells were 0.881 µM and 0.205 µM) and 23g (the
IC50 values against A549 and Hela cells were 0.057 µM and 0.104 µM) both exhibited a excellent dual antiproliferative
activity against A549 and Hela cells.
Table 1 The antiproliferative activities of compounds 21a-g, 22a-g, 23a-g and 24a-g.
Cellular activity IC50 (µM)a
b
Compd.
R1
R2
X
LogD7.4
A549
Hela
21a
21b
21c
21d
21e
21f
F
F
S
S
S
S
S
S
S
S
S
S
S
4.545
0.728
4.0
4.2
4.3
4.6
4.8
5.1
5.2
3.8
4.3
4.2
4.5
0.864
0.902
5.312
1.451
1.154
5.187
0.506
0.716
0.583
6.863
0.980
0.284
0.275
0.839
1.327
0.966
1.297
1.005
2.361
1.648
F
F
F
F
21g
22a
22b
22c
22d
F
H
H
H
H

22e
22f
H
H
H
H
H
H
H
H
H
H
F
S
0.236
0.788
0.571
8.685
0.881
0.787
4.257
1.351
0.038
0.057
4.635
3.183
2.138
3.986
5.832
3.754
0.799
0.711
0.652
0.115
0.205
4.519
6.011
5.224
2.964
0.104
0.175
2.931
5.916
2.471
7.096
0.261
4.6
4.9
5.1
1.8
2.2
2.2
2.5
2.7
2.9
2.6
2.0
2.4
2.4
2.6
2.8
3.1
S
22g
23a
23b
23c
23d
23e
23f
S
SO2
SO2
SO2
SO2
SO2
SO2
SO2
SO2
SO2
SO2
SO2
SO2
SO2
23g
24a
24b
24c
24d
24e
24f
F
F
F
F
F
24g
F
-
SO2
-
7.568
0.462
2.333
0.975
3.3
3.0
AZD9291c
-
a The values are an average of two separate determinations;
b LogD7.4 values were calculated online on the chemaxon website (www.chemaxon.com);
c Used as a positive control.
2.4.2. In vitro EGFR kinase ELISA assay
Based on the results of cellular antiproliferative activities of the target compounds, a partial representative compounds
were selected to test for EGFR kinase activity. The EGFR enzymatic activities of representative compounds were shown
in Table 2. Most of the tested compounds showed potent EGFRT790M kinase inhibitory activity and weak EGFRWT kinase
inhibitory ability. In particular, compounds 22a, 22b, 23b and 23g showed excellent EGFRT790M mutant kinase inhibitory

activity. Their IC50 values over EGFRT790M mutant kinase were 35 nM, 31 nM, 990 nM and 16 nM, respectively. The
compounds with indole ring unsubstituted (22a, 22b and 23g) showed excellent selectivity to EGFRT790M mutant kinase.
Especially the selectivity of compound 23g against EGFRT790M mutant kinase to EGFRWT kinase was more than 125-fold.
The compounds 22a, 22b, 23b and 24c had higher inhibitory activity against EGFRWT kinase with the IC50 values of 511
nM, 570 nM, 104 nM and 201 nM, respectively. It indicated that the acrylamide side chain with small volume structures
substituted were more beneficial to increase the activity of the compounds on EGFRWT kinase, while large volume
substituted groups were more beneficial to increase the inhibitory activity of the compound on EGFRT790M kinase.
Consequently, the results of cellular antiproliferative activity and kinase activity indicated that compounds 22b and 23g
had the excellent dual antiproliferative activity of A549 and Hela cells and EGFRT790M mutant kinase inhibitory ability,
which can be used as potential EGFR inhibitors for further research.
Table 2 The EGFR enzymatic activities of representative compounds.
EGFR IC50 (nM)a
Selectivity
Compd.
WTb
TLc
(WT/TL)
21e
21f
>2000
>2000
511
1128
1689
35
>1.8
>1.2
14.6
18.4
>1.6
>1.2
1.1
22a
22b
570
31
22e
>2000
>2000
104
1248
1621
990
1753
16
22f
23b
23f
>2000
>2000
201
>1.2
>125
-
23g
24c
AZD9291d
>2000
8
16
2.0
a Enzymatic assays were performed by using the ELISA-based EGFR-tyrosine kinase (TK) assay;
b Wide type EGFR kinase;
c T790M and L858R double mutation EGFR kinase;
d Used as a positive control.
2.4.3 In vitro antiproliferative activity against H1975 cells
The results of EGFR enzymatic activities of representative compounds inspired us to further investigate the inhibitory

activity of the target compounds on H1975 cells (EGFRT790M/L858R double mutation). The in vitro inhibition rates of
H1975 cells at 10 uM concentration of representative compound were first conducted to explore the the inhibitory ability
against H1975 cells. As shown in Table 3, all the representative compounds inhibit H1975 cell growth well. Except for
compound 22a, the inhibition rates of other compounds were over 99%. The results indicated that the representative
compounds almost completely inhibited the growth of H1975 cells at a concentration of 10 µM. Next, we tested the IC50
value of compound 22b, 23b and 23g in detail with the AZD9291 as a positive control. The most representative
compound 23g showed potent inhibitory ability over H1975 cells with the IC50 value of 0.916 µM. The antiproliferation
activity results of H1975 cells further indicate that the thiopyranopyrimidine derivative have multiple cell inhibitory
activities, which can be used not only as EGFRT790M/L858R double mutation inhibitors, but also as potential cervical
cancer inhibitors.
Table 3 The antiproliferative activities of compounds 21f, 22a, 22b, 23b, 23g and 24f.
H1975 Cellular activity
Compd.
Inhibitiona
IC50 (µM)b
21f
22a
99.68 %
55.61 %
99.56 %
99.05 %
99.45 %
99.44 %
-
-
-
22b
4.211
4.230
0.916
-
23b
23g
24f
AZD9291c
0.073
a Inhibition rate of H1975 cells at 10 uM concentration;
b The values are an average of two separate determinations;
c Used as a positive control;
- Means not conduct experiment.
2.4.4 In silico molecular docking study
The molecular docking studies were performed to evaluate the structure-activity relationship between different
thiopyranopyrimidine derivatives at the molecular level. As shown in Figure 4, compounds 22b, 23b, 23g and 24c were
well embedded in the hydrophobic cavity of the EGFRT790M protein and formed multiple hydrogen bonds with the
EGFRT790M protein. The thiopyranopyrimidine structure of compound 22b formed a bidentate hydrogen bond with the
hinge region residue Met793 of the EGFRT790M protein with the hydrogen bond lengths of 1.9
Å and 2.6 Å, respectively.
The thiopyran skeleton was far away from the Met790 mutant residue, indicating that the T790M mutation will not
impact the combination of the compounds and the EGFRT790M protein. Furthermore, the O-atom of the acrylamide side
chain also formed a hydrogen bond with the conservative region residue Cys797. The hydrogen bond length was 1.8 Å.

The binding modes of compounds 23b and 24c to EGFRT790M protein were similar to the compound 22c, and the
difference was that the ethanediamine side chain of compound 23b formed a weaker hydrogen bond with residue Leu718
with the hydrogen bond length of 13.1 Å. However, the hydrogen bonding force of compounds 23b and 24c was weaker
than the compound 22c. In particular, the hydrogen bond length between compound 24c and hinge region residue Met793
were only 2.0
Å and 2.5 Å, respectively, and the length of hydrogen bond formed with Cys797 was only 2.2 Å, which was
obviously weaker than the other compounds (22b, 23b and 23g). The compound 23g had the strongest binding ability to
EGFRT790M protein. The length of the hydrogen bond formed with residue Met793 was only 1.7
and 2.4 Å, respectively,
Å
and the length of the hydrogen bond formed with residue Cys797 was only 1.6 Å. The results of molecular docking study
further indicated that the binding modes of compounds 22b and 23g to EGFRT790M protein were similar to those of the
leading compound AZD9291, and the binding abilities were stronger.
B
A
C
D
Figure 4 The docking results of potent compounds bound to EGFRT790M protein (PDB code:3IKA). (A: compound 22b,
B: compound 23b, C: compound 23g, D: compound 24c)
2.4.5 Effect on apoptosis
The promising antiproliferative activity of compounds 22b and 23g inspired us to further investigate the growth
inhibition mechanism of tumor cells. After A549 cells were treated with potent compounds 22b and 23g for 48 h, the
flow cytometric apoptosis experiments were carried out. The results were shown in Figure 5. Compared with the blank
group (Figure 5A, the total apoptosis rate was 5.25%), the lead compound AZD9291 (Figure 5B) significantly induced
A549 cells apoptosis at the concentration of 0.462 µM, and the total apoptosis rate was 15.24%. The same results were
found in compounds 22b and 23g. The apoptosis occurred in A549 cells when treated with 0.716 µM compounds 22b
and 0.057 µM compounds 23g, respectively. The total apoptosis rates of compounds 22b and 23g were 13.55% and

11.79%, respectively. The results of apoptosis indicated that the target compounds 22b and 23g significantly induced
apoptosis. The inhibition effect was similar to that of the lead compound AZD9291.
Figure 5 Cell cycle progression analyses of A549 cells treated with target compounds for 24 h. (A: blank group, B: 0.462
µM AZD9291, C: 0.716 µM compound 22b, D: 0.057 µM compound 23g)
2.4.6 Effect on cell cycle progression
To further investigate the mode of cell proliferation inhibition by compound 22b and 23g, cell cycle distribution analyses
were implemented on A549 cancel cells. The results were presented in Figure 6. Except for the blank group, the
experimental concentrations of all compounds were the IC50 values for A549 cells (See Table 1 for detailed data). As
shown in Figure 6A, the G2/M phase population of the blank group was 3.45%. After treatment with AZD921(Figure
6B), compound 22b (Figure 6C) and compound 23g(Figure 6D), the G2/M phase population of the A549 cancer cells
were all increased, with the G2/M phase population of 14.29%, 12.35% and 10.63%, respectively. However, the S phase
not found significant changes. The results indicated that compound 22b and compound 23g successfully inhibited A549
cells proliferation at a low concentration.
G0/G1=74.31%
S=22.24%
G0/G1=73.81%
S=11.90%
G0/G1=60.51%
S=27.14%
G0/G1=69.17%
S=20.20%
G2/M=3.45%
Blank
G2/M=14.29%
0.462 μM
G2/M=12.35%
0.716 μM
G2/M=10.63%
0.057 μM
Figure 6 Apoptosis analyses of A549 cells treated with target compounds for 24 h. (A: blank group, B: 0.462 µM
AZD9291, C: 0.716 µM compound 22b, D: 0.057 µM compound 23g)
2.4.7 Effect on EGFR signal pathway in Hela cells
To further investigate the effect of the target compound on EGFR signal pathway in Hela cells, the Western Blot assays
were carried out. We tested the inhibitory effect of potent compound 23g on EGFR signal pathway in Hela cells for 24h
with the AZD9291 as a positive control. The western blot assays results (Figure 7) showed both AZD9291 and

compound 23g could inhibit the expression of EGFR protein in Hela cells, but the inhibitory effect on phosphorylation of
EGFR protein was not obvious. The expression of EGFR protein was significantly inhibited as the concentration of
compound 23g increased. In addition, we also observed that the inhibitory effect of AZD9291 was twice as potent as that
of compound 23g, which was consistent with their kinase activity test results in Table 2. The kinase activity test results
and western blot assays results further showed that compound 23g is a potent EGFR inhibitor
.
Figure 7 Effect on EGFR signal pathway in Hela cells for 24h.
2.5. 3D-QSAR study
2.5.1 3D-QSAR model
All structures of thianopyrimidine derivatives were constructed and aligned onto the common substructure
2-((3-aminophenyl)amino)-7,8-dihydro-5H-thiopyrano[4,3-d]pyrimidine 6,6-dioxide nucleus, with the potent compound
23g as template (Figure 8A). As shown in Figure 8B, the structures of all compounds were aligned together. In order to
further study the SARs of thianopyrimidine derivatives, the in vitro anti-proliferative activity data of Hela cells were
chose to construct the 3D-Quantitative structure activity relationship (3D-QSAR) models. Based on the detailed
operation manual reported in the literature29-31, the IC50 values of all compounds against Hela cells were chose and
converted to pIC50 values.
A
B
Figure 8 (A) Structure of the template compound 23g, (B) 3D-QSAR structure alignment and superposition of all
compounds using compound 23g as the template.
2.5.2 CoMFA and CoMSIA analysis

The detailed PLS statistics results of the CoMFA and CoMSIA models were showed in Table 3. The experimental and
the predicted pIC50 values of the target compounds against Hela cancer cells are listed in Table 5. Furthermore, the
graphical relationships of CoMFA and CoMSIA models are illustrated in Figure 9. The cross-validation correlation
coefficient (q2) and correlation coefficient (r2) are two key parameters for evaluating the quality of PLS analysis, which
represented the predictive ability and the self-consistence of the model. A q2 value of more than 0.3 will be considered
statistically as the chance of significant correlation being <95%. The detailed PLS data were shown in Table 3. The q2
and r2 of the CoMFA model were 0.765 and 0.965, respectively, and the q2 and r2 of the CoMSIA model were 0.875 and
0.956, respectively. The statistical results revealed that the models of CoMFA and CoMSIA had reliable predictive ability.
The optimal numbers of components used to generate CoMFA and CoMSIA models was 8 and 5, respectively. It
indicated that the number of compounds used to derive the models were reasonable. The Fisher test results of CoMFA
and CoMSIA models were 65.657 and 94.497, respectively. In addition, the estimated standard errors of the CoMFA and
CoMSIA models were also reasonably low, reaching to 0.123 and 0.129, respectively. As illustrated in Table 4, the
contribution of the electrostatic and steric fields of the CoMFA model was 51.6% and 48.4%, respectively. The
contribution of the steric, electrostatic, hydrophobic, hydrogen bond donor and acceptor fields of the CoMSIA model
were 5.0%, 44.3%, 33.0%, 15.5% and 2.2%, respectively. The PLS data revealed that the CoMFA and CoMSIA models
established are reliable and can be used to guide researchers to carry out in-depth and reasonable modification of the
target molecules for obtaining better activity.
Table 4 PLS statistics of CoMFA and CoMSIA 3D-QSAR models.
q2 a
N b
r2 c
SE d
F e
Fractionf
Steric
Electrostatic
0.516
Hydrophobic Donor
Acceptor
–
CoMFA
0.765
0.875
8
5
0.965 0.123
0.956 0.129
65.657 0.484
94.497 0.050
–
–
CoMSIA
0.443
0.330
0.155
0.022
a
Cross-validated correlation coefficient;
b
Optimum number of components obtained from cross-validated PLS analysis and same used in final non-cross-validated
analysis;
c
d
e
Non-cross-validated correlation coefficient;
Standard error of estimate;
F-test value. f Field contributions.
Table 5 Experimental and predicted pIC50 values against Hela cell line of compounds for CoMFA and CoMSIA models.
CoMFA
CoMSIA
Compd.
pIC50(experimental)
pIC50(predicted)
Residuals
pIC50(predicted)
Residuals
9.126
9.058
9.558
9.594
0.0166
-0.05
9.092
9.124
9.415
9.458
0.0462
0.116
21a
21b
21c
21d
9.138
9.008
9.547
9.561
-0.011
-0.0329
0.1319
0.1032

9.063
8.850
8.975
9.000
8.940
8.661
8.750
9.098
9.059
9.337
9.798
9.784
8.233
8.447
8.368
8.359
9.981
9.883
8.666
8.129
8.503
8.311
9.649
8.595
0.0129
0.0275
0.0400
-0.1134
0.0581
-0.0342
0.0329
-0.0013
0.0886
-0.1514
0.1409
0.096
9.115
8.976
8.944
9.030
9.066
8.592
8.891
9.055
9.006
9.288
9.780
9.844
8.267
8.322
8.353
8.303
10.049
9.820
8.540
8.291
8.400
8.395
9.741
8.622
-0.0388
-0.0993
0.0710
-0.1432
-0.0682
0.0353
-0.1080
0.0418
0.1420
-0.1022
0.1592
0.156
21e
21f
9.076
8.877
9.015
8.887
8.998
8.627
8.783
9.097
9.148
9.186
9.939
9.688
8.345
8.221
8.282
8.528
9.984
9.758
8.533
8.228
8.607
8.149
9.583
8.632
21g
22a
22b
22c
22d
22e
22f
22g
23a
23b
23c
23d
23e
23f
0.1121
-0.2264
-0.0857
0.1694
0.0027
-0.1253
-0.1332
0.0989
0.1042
-0.1621
-0.0655
0.0372
0.0782
-0.1007
-0.0706
0.2246
-0.0646
-0.0616
-0.0073
-0.0626
0.2070
-0.2461
-0.1576
0.0095
23g
24a
24b
24c
24d
24e
24f
24g
The linear correlation results in Figure 9 revealed that the calculated pIC50 and the experimental pIC50 values of the
CoMFA and CoMSIA models had a good linear relationship. In addition, due to structural differences, a small
amplitude fluctuations were observed between the calculated pIC50 and the experimental pIC50 values.
B
A
Figure 9 Calculated pIC50 versus experimental pIC50 values for the target compounds obtained by PLS analysis using
CoMFA (Figure 9A) and CoMSIA (Figure 9B) models.
2.5.3 Countour analysis

In order to better reveal the effect of various fields on the target property, by Sybyl 8.1 software, CoMSIA and COMFA
contour maps were generated to identify the effects of different fields (the steric field, electrostatic field, hydrophobic
field, hydrogen bond donor and acceptor fields) on the activity data of compounds. StDev*Coeff mapping option was
used to convert CoMFA and CoMSIA models into visual results. All contour maps represented 80% level contributions
for favored and 20% level contributions for disfavored. As showed in Figure 10 and Figure 11, the template compound
23g (with the highest pIC50 value) and the compound 21b were selected to better reveal the 3D-QSAR information of
CoMFA and CoMSIA models. The results can better reveal the relationship between structure and biological activity.
A
B
Figure 10 CoMFA contour maps of target compounds. (A: The CoMFA model of compound 21b, B: The CoMFA model
of compound 23g. CoMFA steric contours: the green contours indicate the areas that are conducive to steric interaction,
the yellow contours indicate the areas that are not conducive to steric interaction. CoMSIA electrostatic contours: the
blue contours indicate the regions that are favorable to the positively charged groups, the red contours indicate the
regions that are favorable to the negatively charged groups32.)
A
B
C
D
Figure 11 CoMSIA contour maps of representative compound 23g. (A: The CoMSIA hydrogen bond donor and acceptor

contours: the cyan and the purple contours represent favorable and unfavorable hydrogen bond donor regions,
respectively. The magenta and the red contours represent favorable and unfavorable hydrogen bond acceptor regions,
respectively. B: The CoMSIA electrostatic contours: the blue contour indicates electropositive charge and the red
contour electronegative charge. C: The CoMSIA hydrophobic contours: the yellow contours represent regions where
hydrophobic groups increase activity, while the white contours highlight regions that would favor hydrophilic groups. D:
The CoMSIA steric contours: the green contour favors steric or bulky group and the yellow contour denotes disfavored
region32.)
2.5.4 CoMFA contour analysis.
According to the COMFA contour maps shown in Figure 10, it can be clearly seen that the thiopyran ring was covered
with a green contour. The results indicated that increasing the volume of this site (pyrimidine ring) was beneficial to
improving the antiproliferative activity of the target compound over Hela cells . Therefore, it is reasonable to introduce
the thiopyran structure on the pyrimidine ring. In addition, the result also illustrated the importance of the ring expansion
strategy in drug design. The benzene ring structure of the indole ring was covered with a red contour, denoting that the
introduction of an electronegative group into indole ring was beneficial to enhance the activity. As shown in Figure 10,
compound 21a-g (indole ring with fluorine substitution) had higher anti-proliferative activity of Hela cells than
compound 23a-g (indole ring without fluorine substitution). The results were consistent with the predictions of the
CoMFA model. A big red contours appeared around the N-atom of pyrimidine ring and the S-atom of thiopyran ring,
respectively, indicating that the thiopyranopyrimidine structure played a key role in maintaining the antiproliferation
activity of the target compounds. The appearance of the yellow contours around the acrylamide side chain further
explained why the antiproliferative activity of the target compound decreased as the volume of the acrylamide side chain
increased. In brief, the CoMFA results are consistent with the antiproliferation activity results of Hela cells
(21c>21e>21g, 23b>23c>23e) obtained from our experiments.
2.5.5 CoMSIA contour analysis
The steric contours and electrostatic contours of CoMSIA models were similar to the CoMFA models, but there are some
differences between CoMSIA and CoMSIA models. In the CoMSIA model, we observed that the blue contours (Figure
11B) and the yellow contours (Figure 11D) appeared around the N-methylindole ring, indicating that this region would
be not suitable for bulkier and more electronegative substituents.
The differences between the CoMSIA model and the CoMFA model were that the CoMSIA model can provide
additional hydrogen b006Fnd donor and acceptor contours and hydrophobic contours. Figure 11A showed the hydrogen
bond donor and acceptor contours of the CoMSIA model. A big magenta contour appeared around the S-atom of the

thiopyran ring indicated that the thiopyran ring plays a key role in maintaining the antiproliferative activity of the target
compounds. The Figure 11C showed that a large number of yellow contours overlapped around the acrylamide side
chain. It revealed that the alkylation substitutions of the acrylmide side chain is beneficial to increase the antiproliferative
activity against Hela cells. Similarly, the results of CoMSIA hydrophobic contours are also consistent with the
antiproliferation activity results of Hela cells (21a>21b, 22a>22b, 23a>23b and 24a>24b). In addition, the appearance
of a small amount of yellow contours around the indole ring indicated that the hydrophobicity of the indole ring was
conducive to maintaining the proliferative activity of the compounds against Hela cells.
3. Conclusions
In summary, the structure-based drug design strategy was used to design and synthesize four series of
thiopyranopyrimidine AZD9291 derivatives containing acrylamide structure. First, the structure of acrylamide warhead
was modified to regulate the reactivity of Michael covalent addition so that the affinity of the medicine to non-target
proteins were reduced. Next, the ring expansion principle was applied to expend the volume of the pyrimidine ring. By
this way, the affinity of compounds to EGFR protein can be increased. Finally, fluorine element was also introduced into
the indole ring to get the optimized compound 23g with the IC50 values of 0.057 µM and 0.104 µM against A549 cells
and Hela cells, respectively. The 3D-QSAR results suggested that the introduction of thiopyran skeleton was favorable
for increasing the antiproliferative activity of target compounds against Hela cells. Furthermore, the introduction of bulk
alkyl structure on the acrylamide side chain could significantly increase the antiproliferative activity of the target
compounds against Hela cells. The data of CoMFA (q2=0.765, r2=0.965) and CoMSIA (q2=0.875, r2=0.956) models were
statistically reliable and could be used in the design of novel and potent double antitumor agents. The further studies on
the mechanism of action of these compounds remain in progress and will be published in the future.
4. Experimental section
4.1. General methods of chemistry
Unless otherwise required, all solvent and chemicals purchased were analytical grade and used without further
purification. Frequently used solvents (DCM, PE and EA, etc.) were resteamed to remove water. All chemical reactions
were monitored through GF254 thin-layer chromatography plate and spots were visualized with iodine or light (in 254 nm
or 365 nm). The structure of the target compound was confirmed by 1H NMR and 13C NMR spectra at room temperature
on Bruker 400 MHz spectrometer (Bruker Bioscience, Billerica, MA, USA) with tetramethylsilane (TMS) as an internal
standard. Mass spectrometry (MS) was performed on Waters High Resolution Quadrupole Time of Flight Tandem Mass
Spectrometry (QTOF). The purity of the compound was determined by Agilent 1260 liquid chromatograph fitted with an

Inertex-C18 column. All target compounds had a purity of ≥95%.
4.1.1. General procedure for the synthesis of 21a-g
Commercially available starting material 5-fluoroindole 7b reacted with iodomethane to get 8b. 10 was obtained by the
reaction of self-ester condensation from commercially available 9. Next, cyclization 10 with urea to obtain 11. Then, 11
reacted with phosphorus oxychloride to obtain intermediate 12. Intermediate 12 and 8b undergone Friedel-Crafts
alkylation reaction to obtain intermediate 13b. 14b were obtained by the reaction of oxidation reduction from 13b. The
detailed preparation proccess of the target compounds can be found in our previous research33,34 or the patent published
by AstraZeneca35.
4.1.1.1.
N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-7,8-dihydro-5H-
thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl)acrylamide (21a)
Light grayish powder. 48.3% yield, m.p: 268.7-269.2 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.72 (s, 1H, -CO-NH-), 8.60
(s, 1H, -NH-), 8.06–7.93 (m, 2H, indole-H, Ar-H), 7.78 (s, 1H, indole-H), 7.44 (s, 1H, indole-H), 6.98 (s, 2H, indole-H,
Ar-H), 6.91 (s, 1H, -CH2-CH-), 6.12 (d, J = 16.4 Hz, 1H, -CH=CH2), 5.61 (d, J = 8.2 Hz, 1H, -CH=CH2), 3.85 (s, 2H,
-S-CH2-), 3.82 (s, 3H, -OCH3), 3.78 (s, 3H, indole-CH3), 3.12 (s, 4H, -S-CH2-CH2-, -N-CH3-CH2-), 2.95 (s, 5H, -S-CH2-,
-N-CH3), 2.57 (d, J = 11.8 Hz, 8H, N-(CH3)2, -CH2-). TOF MS ES+ (m/z): (M + H)+, calcd for C28H32FN7O2: 590.2713,
found, 590.2735.
4.1.1.2.
2-chloro-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-7,8-
dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl) acrylamide (21b)
Light grayish powder. 49.7% yield, m.p: 273.2-274.7 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.84 (s, 1H, -CO-NH-), 8.75
(s, 1H, -NH-), 8.07 (s, 1H, indole-H), 8.03 (s, 1H, Ar-H), 7.80 (d, J = 10.4 Hz, 1H, indole-H), 7.50–7.46 (m, 1H,
indole-H), 7.04 (t, J = 9.2, 2.6 Hz, 1H, indole-H), 6.99 (s, 1H, Ar-H), 6.82–6.73 (m, 1H, -C=CH2), 6.03 (d, J = 15.3 Hz,
1H, -C=CH2), 3.90 (s, 2H, -S-CH2-), 3.87 (s, 3H, -OCH3), 3.80 (s, 3H, indole-CH3), 3.00 (dt, J = 10.2, 5.2 Hz, 6H,
N-(CH3)2), 2.89 (s, 3H, -N-CH3), 2.68 (s, 4H, -S-CH2-, -N-CH3-CH2-), 2.36 (s, 2H, -S-CH2-CH2-), 2.21–2.15 (m, 2H,
-CH2-). 13C NMR (400 MHz, DMSO-d6) δ 164.07, 161.95, 160.28, 157.20, 155.46, 147.19, 139.48, 136.09, 135.86,
133.25, 127.59, 127.29, 126.84, 122.84, 115.06, 113.65, 111.94, 111.81, 108.36, 107.97, 97.79, 57.76, 56.15, 52.17,
45.74, 45.71, 38.75, 32.76, 32.33, 30.67, 28.02. TOF MS ES+ (m/z): (M + H)+, calcd for C28H32FN7O2: 624.2324, found,
624.2313.
4.1.1.3.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-7,8-dihydro-5H-
thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl)but-2-enamide (21c)

Light grayish solid. 50.9% yield, m.p: 271.3-273.8 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.86 (s, 1H, -CO-NH-), 8.76 (s,
1H, -NH-), 8.07 (s, 1H, indole-H), 8.03 (d, J = 3.2 Hz, 1H, Ar-H), 7.48 (dd, J = 8.6, 4.6 Hz, 1H, indole-H), 7.03 (s, 1H,
indole-H), 7.00 (d, J = 3.4 Hz, 1H, indole-H), 6.72 (dd, J = 15.3, 6.3 Hz, 1H, Ar-H), 5.99 (d, J = 15.3 Hz, 1H, -CH=C-),
3.90 (s, 2H, -S-CH2-), 3.87 (s, 3H, -OCH3), 3.81 (s, 3H, indole-CH3), 3.05 – 2.97 (m, 5H, -N-CH3, -N-CH3-CH2-), 2.89 (s,
2H, -S-CH2-), 2.68 (s, 4H, -CH2-, -S-CH2-CH2-), 2.50 (s, 6H, N-(CH3)2), 2.28 (s, 6H, C-(CH3)2). 13C NMR (400 MHz,
DMSO-d6) δ 166.22, 161.95, 160.28, 157.20, 155.46, 153.52, 147.07, 139.73, 134.92, 133.19, 128.26, 127.25, 126.84,
115.31, 115.00, 113.65, 111.88, 111.80, 109.40, 108.13, 97.85, 57.71, 56.15, 52.20, 45.74, 45.71, 38.72, 32.76, 32.31,
30.67, 28.02, 25.32, 20.70.
4.1.1.4.
N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-7,8-dihydro-5H-
thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl)-3-methylbut-2-enamide (21d)
Light grayish solid. 47.8% yield, m.p: 274.8-276.1ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.72 (s, 1H, -CO-NH-), 8.69 (s,
1H, -NH-), 8.06 (s, 1H, indole-H), 8.03 (s, 1H, Ar-H), 7.83–7.78 (m, 1H, indole-H), 7.48 (dd, J = 9.0, 4.5 Hz, 1H,
indole-H), 7.06–7.00 (m, 1H, indole-H), 6.98 (s, 1H, Ar-H), 6.70 (dt, J = 13.9, 7.0 Hz, 1H, -CH-CH3), 6.17 (s, 1H,
-CO-CH-), 3.90 (s, 2H, -S-CH2-), 3.86 (s, 3H, -OCH3), 3.81 (s, 3H, indole-CH3), 3.03–2.97 (m, 6H, -S-CH2-CH2-,
-N-CH3-CH2-), 2.65 (s, 5H, -N-CH3, -CH2-), 2.50 (s, 3H, -CH-CH3), 2.36 (s, 6H, N-(CH3)2). TOF MS ES+ (m/z): (M +
H)+, calcd for C28H32FN7O2: 618.3026, found, 618.3108.
4.1.1.5.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-7,8-dihydro-5H-
thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl)pent-2-enamide (21e)
Light grayish solid. 47.2% yield, m.p: 273.8-275.2 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.76 (s, 1H, -CO-NH-), 8.75 (s,
1H, -NH-), 8.10 (s, 1H, indole-H), 8.07 (s, 1H, Ar-H), 7.84 (d, J = 11.2 Hz, 1H, indole-H), 7.51(dd, J = 8.5, 4.8 Hz, 1H,
indole-H), 7.07–7.03 (m, 1H, indole-H), 6.95 (d, J = 5.1 Hz, 1H, Ar-H), 6.77 (dt, J = 13.2, 6.3 Hz, 1H, -CO-CH=CH-),
6.15 (s, 1H, -CO-CH=), 3.92 (s, 2H, -S-CH2-), 3.89(s, 3H, -OCH3), 3.82 (s, 3H, indole-CH3), 3.09 (dd, J = 7.7, 4.2 Hz,
5H, -N-CH3, -S-CH2-), 2.68 (s, 4H, -CH2-, -S-CH2-CH2-), 2.35 (s, 6H, N-(CH3)2), 2.14 (q, J = 6.3 Hz, 4H, -CH2-CH3,
-N-CH3-CH2-), 1.01 (t, J = 7.3 Hz, 3H, -CH2-CH3). 13C NMR (400 MHz, DMSO-d6) δ 165.14, 161.95, 158.27, 157.20,
155.42, 147.84, 147.07, 139.73, 135.03, 133.35, 128.14, 127.25, 126.91, 120.58, 115.09, 113.65, 112.00, 111.80, 109.40,
108.13, 97.79, 57.85, 56.15, 51.77, 45.74, 45.44, 38.77, 32.73, 32.17, 30.69, 28.09, 22.95, 12.08.
4.1.1.6.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-7,8-dihydro-
5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl)-4-methylpent-2-enamide (21f)
Light grayish solid. 51.8% yield, m.p: 278.4-279.6 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.87 (s, 1H, -CO-NH-), 8.76 (s,
1H, -NH-), 8.07 (s, 1H, indole-H), 8.03 (s, 1H, Ar-H), 7.79 (d, J = 11.0 Hz, 1H, indole-H), 7.48 (dd, J = 8.9, 4.6 Hz, 1H

indole-H), 7.04 (d, J = 8.7 Hz, 1H, indole-H), 6.99 (s, 1H, Ar-H), 6.72 (dd, J = 15.1, 6.4 Hz, 1H, -CO-CH=), 5.99 (d, J =
16.1 Hz, 1H, -CO-CH=CH-), 3.90 (s, 2H, -S-CH2-), 3.87 (s, 3H, -OCH3), 3.81 (s, 3H, indole-CH3), 3.05–2.97 (m, 4H,
-N-CH3-CH2-, -S-CH2-), 2.89 (s, 2H, -S-CH2-CH2-), 2.68 (s, 3H, -N-CH3), 2.45 (s, 2H, -CH2-), 2.28 (s, 6H, N-(CH3)2),
2.18 (t, J = 7.5 Hz, 1H, CH-(CH3)2), 1.05 (d, J = 6.8 Hz, 6H, CH-(CH3)2). 13C NMR (400 MHz, DMSO-d6) δ 164.50,
161.95, 158.25, 157.20, 155.46, 147.07, 146.71, 139.55, 135.04, 133.25, 128.15, 127.25, 126.84, 121.93, 115.07, 113.65,
112.02, 111.76, 109.36, 108.36, 97.79, 57.79, 56.15, 52.15, 45.74, 45.71, 38.72, 33.04, 32.19, 31.15, 30.38, 27.97, 21.52,
21.18.
4.1.1.7.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-7,8-dihydro-
5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl)hex-2-enamide (21g)
Light grayish solid. 50.4% yield, m.p: 279.6-281.2 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.78 (s, 1H, -CO-NH-), 8.72 (s,
1H, -NH-), 8.07 (s, 1H, indole-H), 8.03 (s, 1H, Ar-H), 7.80 (d, J = 10.7 Hz, 1H, indole-H), 7.48 (dd, J = 8.9, 4.6 Hz, 1H,
indole-H), 7.07–7.01 (m, 1H, indole-H), 6.99 (d, J = 5.5 Hz, 1H, -CO-CH=CH-), 6.71 (dt, J = 14.7, 6.8 Hz, 1H,
-CO-CH=), 6.12 (s, 1H, Ar-H), 3.90 (s, 2H, -S-CH2-), 3.87 (s, 3H, -OCH3), 3.81 (s, 3H, indole-CH3), 3.00 (dd, J = 7.8,
4.4 Hz, 5H, -S-CH2-, -N-CH3), 2.66 (s, 4H, -S-CH2-CH2-, -N-CH3-CH2-), 2.33 (s, 6H, N-(CH3)2), 2.18 (q, J = 6.5 Hz, 4H,
-CH2-CH2-CH3, -CH2-), 1.46 (d, J = 7.3 Hz, 2H, -CH2-CH3), 0.91 (t, J = 7.3 Hz, 3H, -CH2-CH3). TOF MS ES+ (m/z): (M
+ H)+, calcd for C28H32FN7O2: 632.3183, found, 632.3195.
4.1.2. General procedure for the synthesis of 22a-g
The synthetic methods of compounds 22a-g are similarly to 21a-g. The difference is that we need to replace the
5-fluoroindole with indole and the rest of the synthesis conditions are the same.
4.1.2.1.
N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-7,8-dihydro-5H-t
hiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)acrylamide (22a)
Light yellow solid. 51.2% yield, m.p: 256.8-258.1 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.90 (s, 1H, -CO-NH-), 9.20 (s,
1H, -NH-), 8.40 (s, 1H, indole-H), 8.34 (s, 1H, indole-H), 8.03 (s, 1H, indole-H), 7.51 (d, J = 6.8 Hz, 1H, Ar-H), 7.23 (t,
J = 7.0 Hz, 2H, indole-H), 7.09 (t, J = 8.0 Hz, 1H, Ar-H), 7.00 (s, 1H, -CH=CH2), 6.18 (d, J = 17.1 Hz, 1H, -CH=CH2),
5.66 (d, J = 10.9 Hz, 1H, -CH=-CH2), 3.97 (s, 2H, -S-CH2-), 3.89 (s, 3H, -OCH3), 3.83 (s, 3H, indole-CH3), 3.33 (s, 2H,
-N-CH3-CH2-), 3.13 (s, 2H, -S-CH2-), 3.00 (s, 2H, -S-CH2-CH2-), 2.71 (d, J = 14.2 Hz, 8H, -CH2-, N-(CH3)2), 2.63 (s, 3H,
-N-CH3). 13C NMR (400 MHz, DMSO-d6) δ 164.73, 161.95, 157.20, 154.43, 147.80, 139.70, 136.39, 132.41, 128.24,
127.84, 127.18, 127.16, 126.03, 122.99, 121.92, 120.08, 115.66, 113.62, 110.18, 109.38, 97.81, 57.71, 56.15, 52.20,
45.78, 45.74, 38.72, 32.73, 32.34, 30.63, 28.02.
4.1.2.2.
2-chloro-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-7,8-

dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)acrylamide (22b)
Light yellow solid. 52.4% yield, m.p: 258.6-259.9 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.80 (s, 1H, -CO-NH-), 8.67 (s,
1H, -NH-), 8.10–8.05 (m, 2H, indole-H), 7.51 (d, J = 8.1 Hz, 1H, indole-H), 7.22 (s, 1H, Ar-H), 7.11 (d, J = 9.2 Hz, 2H,
indole-H), 7.04 (s, 1H, Ar-H), 6.52 (s, 1H, -C=CH2), 6.08 (s, 1H, -C=CH2), 3.94 (s, 2H, -S-CH2-), 3.90 (s, 3H, -OCH3),
3.88 (s, 3H, indole-CH3), 3.17 (s, 2H, -N-CH3-CH2-), 3.07 (s, 2H, -S-CH2-), 3.02–3.00 (m, 2H, -S-CH2-CH2-), 2.77 (s,
6H, N-(CH3)2), 2.73 (s, 2H, -CH2-), 2.63 (s, 3H, -N-CH3). 13C NMR (400 MHz, DMSO-d6) δ 164.07, 161.95, 157.20,
154.38, 147.90, 139.48, 136.64, 135.86, 132.49, 127.59, 127.29, 126.03, 122.84, 122.42, 121.98, 120.14, 115.66, 113.64,
110.18, 108.51, 97.66, 57.79, 56.15, 52.15, 45.76, 45.72, 38.75, 32.73, 32.31, 30.64, 28.02. TOF MS ES+ (m/z): (M +
H)+, calcd for C28H32FN7O2: 606.2418, found, 606.2411.
4.1.2.3.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-7,8-dihydro-
5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)but-2-enamide (22c)
Light yellow solid. 47.9% yield, m.p: 257.2-259.3 ºC. 1H NMR (400 MHz, DMSO-d6) δ 10.22 (s, 1H, -CO-NH-), 9.11 (s,
1H, -NH-), 8.11 (d, J = 8.0 Hz, 1H, indole-H), 7.85 (dd, J = 8.2, 4.3 Hz, 1H, indole-H), 7.68 (d, J = 8.4 Hz, 1H, Ar-H),
7.25 (t, J = 7.7 Hz, 1H, indole-H), 7.18 (s, 1H, indole-H), 6.88 (s, 1H, indole-H), 6.47 (s, 1H, Ar-H), 6.09 (s, 1H,
-CO-CH), 3.87 (s, 3H, -OCH3), 3.72 (s, 3H, indole-CH3), 3.68 (s, 2H, -S-CH2-), 3.36 (s, 4H, -N-CH3-CH2-, -S-CH2-),
3.20 (d, J = 5.9 Hz, 2H, -S-CH2-CH2-), 2.96 (s, 3H, -N-CH3), 2.72 (s, 2H, -CH2-), 2.65 (s, 6H, N-(CH3)2), 2.34 (s, 3H,
-CH-(CH3)2), 2.21 (s, 3H, -CH-(CH3)2). 13C NMR (400 MHz, DMSO-d6) δ 166.15, 161.95, 157.20, 154.43, 153.52,
147.80, 139.74, 136.52, 132.51, 128.56, 127.17, 126.03, 122.49, 121.89, 120.03, 115.71, 115.39, 113.64, 110.13, 109.38,
97.79, 57.79, 56.15, 52.15, 45.74, 45.71, 38.75, 32.73, 32.34, 30.63, 28.02, 25.32, 20.76.
4.1.2.4.
N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-7,8-dihydro-
5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)-3-methylbut-2-enamide (22d)
Light yellow solid. 47.3% yield, m.p: 256.7-258.2 ºC. 1H NMR (400 MHz, DMSO-d6) δ 10.23 (s, 1H, -CO-NH-), 9.85 (s,
1H, -NH-), 8.67 (s, 1H, indole-H), 8.08 (d, J = 8.0 Hz, 1H, indole-H), 8.03 (s, 1H, Ar-H), 7.62 (dd, J = 8.0, 4.7 Hz, 1H,
indole-H), 7.51 (d, J = 8.1 Hz, 1H, indole-H), 7.22 (t, J = 7.7 Hz, 1H, indole-H), 7.13 (t, J = 7.7 Hz, 1H, =CH-CH3), 6.92
(s, 1H, Ar-H), 6.78 (d, J = 5.9 Hz, 1H, -CO-CH=), 3.95 (s, 2H, -S-CH2-), 3.90 (s, 3H, -OCH3), 3.88 (s, 3H, indole-CH3),
3.35 (s, 4H, -N-CH3-CH2-, -S-CH2-), 3.13 (d, J = 6.8 Hz, 3H, -N-CH3), 3.02 (d, J = 6.3 Hz, 2H, -S-CH2-CH2-), 2.75 (s,
6H, N-(CH3)2), 2.72 (s, 2H, -CH2-), 2.66 (s, 3H, -CH-CH3).TOF MS ES+ (m/z): (M + H)+, calcd for C28H32FN7O2:
600.3121, found, 600.3231.
4.1.2.5.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-7,8-dihydro-
5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)pent-2-enamide (22e)

Light yellow solid. 49.1% yield, m.p: 261.3-1263.4 ºC. 1H NMR (400 MHz, DMSO-d6) δ 10.55 (s, 1H, -CO-NH-), 9.59
(s, 1H, -NH-), 8.64 (s, 1H, indole-H), 8.06 (d, J = 8.0 Hz, 1H, indole-H), 8.02 (s, 1H, Ar-H), 7.54 (dd, J = 8.0, 4.7 Hz, 1H,
indole-H), 7.48 (d, J = 8.3 Hz, 1H, indole-H), 7.19 (t, J = 7.6 Hz, 1H, indole-H), 7.09 (t, J = 7.5 Hz, 1H, =CH-CH2-CH3),
6.92 (s, 1H, Ar-H), 6.78 (d, J = 5.4 Hz, 1H, -CO-CH=), 3.91 (s, 2H, -S-CH2-), 3.87 (s, 3H, -OCH3), 3.85 (s, 3H,
indole-CH3), 3.25 (s, 2H, -N-CH3-CH2-), 3.01 (dd, J = 12.8, 5.3 Hz, 4H, -S-CH2-, -S-CH2-CH2-), 2.74 (s, 2H, -CH2-),
2.71 (s, 6H, N-(CH3)2), 2.57 (s, 3H, -N-CH3), 2.20 (p, J = 6.7 Hz, 2H, -CH2-CH3), 1.05 (t, J = 7.4 Hz, 3H, -CH2-CH3). 13C
NMR (400 MHz, DMSO-d6) δ 165.06, 161.95, 157.20, 154.38, 147.80, 147.72, 139.73, 136.47, 132.45, 128.40, 127.17,
126.09, 122.36, 121.89, 120.54, 119.60, 115.82, 113.64, 110.13, 109.38, 97.64, 57.79, 56.15, 52.20, 46.29, 45.37, 38.72,
32.73, 32.35, 30.63, 27.97, 22.96, 12.22. TOF MS ES+ (m/z): (M + H)+, calcd for C28H32FN7O2: 600.3121, found,
600.3120.
4.1.2.6
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-7,8-dihydro-
5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)-4-methylpent-2-enamide (22f)
Light yellow solid. 42.9% yield, m.p: 260.8-262.4 ºC. 1H NMR (400 MHz, DMSO-d6) δ 10.32 (s, 1H, -CO-NH-), 9.21 (s,
1H, -NH-), 8.08 (d, J = 8.0 Hz, 1H, indole-H), 7.81 (dd, J = 8.0, 4.7 Hz, 1H, indole-H), 7.61 (d, J = 8.3 Hz, 1H, Ar-H),
7.21 (t, J = 7.6 Hz, 1H, indole-H), 7.13 (s, 1H, indole-H), 6.92 (s, 1H, indole-H), 6.51 (s, 1H, Ar-H), 6.21 (d, J= 5.4Hz,
2H, -CO-CH=CH-), 3.97 (s, 3H, -OCH3), 3.80 (s, 3H, indole-CH3), 3.78 (s, 2H, -S-CH2-), 3.40 (s, 4H, -N-CH3-CH2-,
-S-CH2-), 3.22 (d, J = 5.9 Hz, 2H, -S-CH2-CH2-), 3.02 (s, 3H, -N-CH3), 2.75 (s, 6H, N-(CH3)2), 2.72 (s, 2H, -CH2-), 2.68
(s, 1H, -CH-(CH3)2), 1.03 (s, 3H, -CH-(CH3)2), 1.01 (s, 3H, -CH-(CH3)2). 13C NMR (400 MHz, DMSO-d6) δ 164.50,
161.95, 157.20, 154.38, 147.80, 146.75, 139.54, 136.37, 132.51, 128.14, 127.25, 126.09, 122.81, 121.93, 121.34, 119.86,
115.72, 113.64, 110.13, 109.40, 97.85, 57.79, 56.15, 52.20, 45.92, 45.35, 38.61, 32.73, 32.01, 31.10, 30.48, 28.02, 21.52,
21.06.
4.1.2.7
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-7,8-dihydro-
5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)hex-2-enamide (22g)
Light yellow solid. 43.6% yield, m.p: 263.4-265.1 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.84 (s, 1H, -CO-NH-), 8.89 (s,
1H, -NH-), 8.04 (d, J = 8.1 Hz, 1H, indole-H), 7.99 (s, 1H, indole-H), 7.84 (s, 1H, Ar-H), 7.47 (d, J = 8.2 Hz, 1H,
indole-H), 7.18 (t, J = 7.6 Hz, 1H, indole-H), 7.04 (t, J = 7.7 Hz, 1H, indole-H), 6.98 (s, 1H, Ar-H), 6.76 (dt, J = 14.7, 6.9
Hz, 1H, -CO-CH=), 6.07 (d, J = 15.2 Hz, 1H, -CO-CH=CH-), 3.90 (s, 2H, -S-CH2-), 3.87 (s, 3H, -OCH3), 3.82 (s, 3H,
indole-CH3), 3.06–2.97 (m, 4H, -N-CH3-CH2-, -S-CH2-), 2.90 (s, 2H, -S-CH2-CH2-), 2.67 (s, 3H, -N-CH3), 2.42 (s, 2H,
-CH2-), 2.32–2.13 (m, 8H, N-(CH3)2, -CH2-CH2-CH3), 1.47 (dt, J = 14.6, 7.3 Hz, 2H, -CH2-CH2-CH3), 0.92 (t, J = 7.3 Hz,
3H, -CH2-CH2-CH3). TOF MS ES+ (m/z): (M + H)+, calcd for C28H32FN7O2: 614.3277, found, 614.3276.

4.1.3. General procedure for the synthesis of 23a-g
The synthetic methods of compounds 23a-g are similarly to 22a-g. The difference is that we need to replace the
intermediate 13b with intermediate 14a and the rest of the synthesis conditions are the same.
4.1.3.1.
N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-6,6-dioxido-7,8-
dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)acrylamide (23a)
Light yellow solid. 38.9% yield, m.p: 238.7-240.2 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.93 (s, 1H, -CO-NH-), 8.70 (s,
1H, -NH-), 8.19 (s, 1H, indole-H), 8.05 (d, J = 8.2 Hz, 1H, indole-H), 7.96 (s, 1H, Ar-H), 7.49 (d, J = 8.3 Hz, 1H,
indole-H), 7.20 (t, J = 7.6 Hz, 1H, indole-H), 7.06 (t, J = 7.6 Hz, 1H, indole-H), 6.99 (s, 1H, Ar-H), 6.66 (s, 1H,
-CH=CH2), 6.22 (d, J = 16.8 Hz, 1H, -CH=CH2), 5.72 (d, J = 10.4 Hz, 1H, -CH=CH2), 4.48 (s, 2H, -SO2-CH2-), 3.87 (s,
3H, -OCH3), 3.83 (s, 3H, indole-CH3), 3.58 (t, J = 7.0 Hz, 2H, -N-CH3-CH2-), 3.29 (t, J = 7.6 Hz, 4H, -SO2-CH2-CH2-),
3.06 (d, J = 13.8 Hz, 2H, -CH2-), 2.65 (s, 3H, -N-CH3), 2.42 (s, 6H, N-(CH3)2). TOF MS ES+ (m/z): (M + H)+, calcd for
C28H32FN7O2: 604.2706, found, 604.2710.
4.1.3.2.
2-chloro-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-6,6-
dioxido-7,8-dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)acrylamide (23b)
Light yellow solid. 39.7% yield, m.p: 239.8-241.1 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.87 (s, 1H, -CO-NH-), 8.69 (s,
1H, -NH-), 8.14 (s, 1H, indole-H), 8.07 (d, J = 8.1 Hz, 1H, indole-H), 7.96 (s, 1H, Ar-H), 7.50 (d, J = 8.1 Hz, 1H,
indole-H), 7.22 (t, J = 7.5 Hz, 1H, indole-H), 7.11 (t, J = 7.5 Hz, 1H, indole-H), 7.06 (s, 1H, Ar-H), 6.53 (s, 1H, -C=CH2),
6.07 (s, 1H, -C=CH2), 4.50 (s, 2H, -SO2-CH2-), 3.89 (d, J = 2.7 Hz, 6H, -OCH3, indole-CH3), 3.60 (t, J = 6.8 Hz, 2H,
-N-CH3-CH2-), 3.32 (d, J = 6.3 Hz, 4H, -SO2-CH2-CH2-), 3.00 (s, 3H, -N-CH3), 2.63 (s, 6H, N-(CH3)2), 2.53–2.50 (m, 2H,
-CH2-). 13C NMR (400 MHz, DMSO-d6) δ 164.43, 159.93, 157.92, 155.05, 147.19, 139.48, 136.60, 135.53, 132.51,
127.59, 127.37, 126.03, 122.84, 122.53, 121.76, 120.11, 116.24, 110.13, 108.36, 104.97, 97.79, 57.79, 56.15, 53.42,
52.20, 48.79, 45.74, 45.57, 38.75, 32.70, 29.64. TOF MS ES+ (m/z): (M + H)+, calcd for C28H32FN7O2: 638.2316, found,
638.2325.
4.1.3.3.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-6,6-dioxido-
7,8-dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)but-2-enamide (23c)
Light yellow solid. 36.9% yield, m.p: 234.5-236.9 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.69 (s, 1H, -CO-NH-), 8.59 (s,
1H, -NH-), 8.16 (s, 1H, indole-H), 8.11 (d, J = 8.0 Hz, 1H, indole-H), 7.93 (s, 1H, Ar-H), 7.51 (d, J = 8.3 Hz, 1H,
indole-H), 7.19 (t, J = 7.7 Hz, 1H, indole-H), 7.07(t, J = 7.5 Hz, 1H, indole-H), 6.97 (s, 1H, Ar-H), 6.78 (dt, J = 14.6, 6.7
Hz, 1H, -CO-CH=), 4.51 (s, 2H, -SO2-CH2-), 3.86 (s, 3H, -OCH3), 3.88 (s, 3H, indole-CH3), 3.73 (t, J = 6.7 Hz, 3H,
-N-CH3), 3.30 (t, J = 7.6 Hz, 2H, -N-CH3-CH2-), 3.22–3.02 (m, 4H, -SO2-CH2-CH2-), 2.72 (s, 6H, N-(CH3)2), 2.64 (d, J =

13.8 Hz, 2H, -CH2-), 2.54 (s, 3H, =C(CH3)2), 2.31 (s, 3H, =C(CH3)2). 13C NMR (400 MHz, DMSO-d6) δ 166.22, 159.93,
157.92, 155.05, 153.52, 147.07, 139.57, 136.55, 132.51, 128.21, 127.25, 126.03, 122.71, 122.05, 120.08, 116.27, 115.31,
110.05, 109.36, 104.97, 97.85, 57.74, 56.15, 53.42, 52.20, 48.79, 45.74, 45.71, 38.72, 32.73, 29.64, 25.38, 20.76.
4.1.3.4.
N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-6,6-dioxido-7,8-
dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)-3-methylbut-2-enamide (23d)
Light yellow solid. 40.2% yield, m.p: 235.9-237.2 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.63 (s, 1H, -CO-NH-), 8.58 (s,
1H, -NH-), 8.17 (s, 1H, indole-H), 8.03 (d, J = 8.0 Hz, 1H, indole-H), 7.93 (s, 1H, Ar-H), 7.47 (d, J = 8.3 Hz, 1H,
indole-H), 7.18 (t, J = 7.7 Hz, 1H, indole-H), 7.05 (t, J = 7.5 Hz, 1H, indole-H), 6.93 (s, 1H, Ar-H), 6.74 (dt, J = 14.6, 6.7
Hz, 2H, -CO-CH=CH-), 4.46 (s, 2H, -SO2-CH2-), 3.86 (s, 3H, -OCH3), 3.81 (s, 3H, indole-CH3), 3.57 (t, J = 6.7 Hz, 2H,
-N-CH3-CH2-), 3.28 (d, J = 6.5 Hz, 4H, -SO2-CH2-CH2-), 3.22–3.02 (m, 3H, -N-CH3), 2.59 (s, 8H, N-(CH3)2, -CH2-),
1.84 (d, J = 6.0 Hz, 3H, =CH-CH3). TOF MS ES+ (m/z): (M + H)+, calcd for C28H32FN7O2: 632.3019, found, 632.3026.
4.1.3.5.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-6,6-dioxido-
7,8-dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)pent-2-enamide (23e)
Light yellow solid. 33.5% yield, m.p: 237.6-239.2 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.56 (s, 1H, -CO-NH-), 8.53 (s,
1H, -NH-), 8.16 (s, 1H, indole-H), 8.06 (d, J = 7.8 Hz, 1H, indole-H), 7.95 (s, 1H, Ar-H), 7.50 (d, J = 7.8 Hz, 1H,
indole-H), 7.23–7.17 (m, 1H, indole-H), 7.12–7.05 (m, 1H, indole-H), 6.94 (s, 1H, Ar-H), 6.87–6.75 (m, 1H, -CO-CH=),
6.67 (s, 1H, -CO-CH=CH-), 4.49 (s, 2H, -SO2-CH2-), 3.88 (s, 3H, -OCH3), 3.85 (s, 3H, indole-CH3), 3.62–3.55 (m, 2H,
-N-CH3-CH2-), 3.31–3.23 (m, 4H, -SO2-CH2-CH2-), 2.71 (s, 6H, N-(CH3)2), 2.58 (s, 3H, -N-CH3), 2.26–2.14 (m, 2H,
-CH2-), 1.30 (s, 2H, -CH2-CH3), 1.06 (t, J = 7.1 Hz, 3H, -CH2-CH3). 13C NMR (400 MHz, DMSO-d6) δ 165.14, 159.93,
157.92, 155.00, 147.84, 147.07, 139.57, 136.49, 132.51, 128.14, 127.25, 126.03, 122.71, 122.05, 120.58, 120.03, 116.27,
110.13, 109.36, 104.97, 97.85, 57.79, 56.15, 53.42, 52.20, 48.79, 45.74, 45.31, 38.75, 32.70, 29.64, 23.08, 12.29.
4.1.3.6.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-6,6-dioxido-
7,8-dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)-4-methylpent-2-enamide (23f)
Light yellow solid. 32.4% yield, m.p: 241.5-243.7 ºC. 1H NMR (400 MHz, DMSO-d6) δ 10.08 (s, 1H, -CO-NH-), 9.31 (s,
1H, -NH-), 8.37 (d, J = 8.9 Hz, 1H, indole-H), 7.71 (d, J = 10.9 Hz, 1H, indole-H), 7.55 (dd, J = 8.2, 4.7 Hz, 1H
indole-H), 7.44–7.32 (m, 1H, indole-H), 7.08 (s, 1H, Ar-H), 6.95 (s, 1H, indole-H), 6.53 (s, 1H Ar-H), 6.39–6.20 (dd, J =
15.2, 6.2 Hz, 2H, -CO-CH=CH-), 4.49 (s, 2H, -SO2-CH2-), 3.88 (s, 3H, -OCH3), 3.82 (s, 3H, indole-CH3), 3.60 (t, J = 6.6
Hz, 3H, -N-CH3), 3.31 (d, J = 5.3Hz, 6H, -SO2-CH2-CH2-, -N-CH3-CH2-), 2.72 (s, 6H, N-(CH3)2), 2.61 (s, 3H, -CH2-,
-CH-(CH3)2), 1.01 (d, J = 6.3 Hz, 6H, -CH-(CH3)2). 13C NMR (400 MHz, DMSO-d6) δ 164.97, 161.93, 158.92, 154.97,
147.07, 146.74, 139.57, 136.60, 132.51, 128.09, 127.25, 126.09, 122.73, 122.04, 121.87, 120.08, 116.48, 110.13, 109.36,

104.97, 97.80, 57.79, 56.15, 53.39, 52.20, 48.79, 45.74, 45.48, 38.72, 32.73, 31.12, 29.61, 21.54, 21.43.
4.1.3.7.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-4-methoxy-5-((4-(1-methyl-1H-indol-3-yl)-6,6-dioxido-
7,8-dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)phenyl)hex-2-enamide (23g)
Light yellow solid. 34.5% yield, m.p: 243.2-245.7 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.62 (s, 1H, -CO-NH-), 8.58 (s,
1H, -NH-), 8.16 (s, 1H, indole-H), 8.05 (d, J = 8.0 Hz, 1H, indole-H), 7.95 (s, 1H, Ar-H), 7.48 (d, J = 8.0 Hz, 1H,
indole-H), 7.19 (t, J = 7.5 Hz, 1H, indole-H), 7.08 (t, J = 7.5 Hz, 1H, indole-H), 6.93 (s, 1H, Ar-H), 6.75 (s, 2H,
-CO-CH=CH-), 4.48 (s, 2H, -SO2-CH2-), 3.87 (s, 3H, -OCH3), 3.83 (s, 3H, indole-CH3), 3.58 (t, J = 6.6 Hz, 2H,
-N-CH3-CH2-), 3.28 (t, J = 6.5 Hz, 2H, -SO2-CH2-), 3.26–3.16 (m, 2H, -SO2-CH2-CH2-), 2.65 (s, 6H, N-(CH3)2), 2.58 (s,
3H, -N-CH3), 2.16 (q, J = 6.5 Hz, 2H, -CH2-), 1.47 (h, J = 6.3, 5.9 Hz, 2H, -CH2-CH2-CH3), 1.24 (d, J = 8.9 Hz, 2H,
-CH2-CH2-CH3), 0.92 (t, J = 7.2 Hz, 3H, -CH2-CH2-CH3). 13C NMR (400 MHz, DMSO-d6) δ 163.87, 162.00, 161.50,
158.83, 147.87, 143.34, 139.51, 136.70, 133.42, 126.45, 126.07, 125.43, 124.43, 122.11, 121.99, 120.64, 118.14, 110.55,
109.96, 108.97, 104.54, 56.02, 50.54, 46.69, 42.82, 42.44, 33.39, 32.90, 32.14, 31.13, 29.83, 28.95, 21.14, 13.61. TOF
MS ES+ (m/z): (M + H)+, calcd for C28H32FN7O2: 646.3175, found, 646.3196.
4.1.4. General procedure for the synthesis of 24a-g
The synthetic methods of compounds 24a-g are similarly to 23a-g. The difference is that we need to replace the indole
with 5-fluroindole and the rest of the synthesis conditions are the same.
4.1.4.1.
N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-6,6-dioxido-7,8-
dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl)acrylamide (24a)
Light yellow solid. 35.8% yield, m.p: 254.4-256.7 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.74 (s, 1H, -CO-NH-), 9.13 (s,
1H, -NH-), 8.39 (s, 1H, indole-H), 8.24 (d, J = 4.0 Hz, 1H, indole-H), 7.76 (s, 1H, Ar-H), 7.53 (dt, J = 8.9, 4.4 Hz, 1H,
indole-H), 7.42–7.33 (m, 1H, Ar-H), 7.08 (tt, J = 9.0, 4.1 Hz, 1H, indole-H), 6.99 (d, J = 7.6 Hz, 1H, -CH=CH2), 6.15 (d,
J = 16.1 Hz, 1H, -CH=CH2), 5.64 (d, J = 9.2 Hz, 1H, -CH=CH2), 4.56 (s, 2H, -SO2-CH2-), 3.89 (s, 3H, -OCH3), 3.81 (s,
3H, indole-CH3), 3.62 (d, J = 6.6 Hz, 3H, -N-CH3), 3.38 (d, J = 6.5 Hz, 2H, -SO2-CH2-), 3.32 (s, 4H, -SO2-CH2-CH2-,
-N-CH3-CH2-), 2.71 (d, J = 4.5 Hz, 6H, N-(CH3)2), 2.60 (s, 2H, -CH2-). 13C NMR (400 MHz, DMSO-d6) δ 164.82,
159.93, 159.00, 157.92, 155.15, 147.07, 139.74, 135.01, 133.39, 127.96, 127.84, 127.25, 127.16, 126.84, 115.96, 111.88,
111.80, 109.40, 108.13, 104.98, 97.85, 57.71, 56.15, 53.42, 52.20, 48.79, 45.74, 45.71, 38.75, 32.73, 29.64. TOF MS ES+
(m/z): (M + H)+, calcd for C28H32FN7O2: 622.2612, found, 622.2615.
4.1.4.2.
2-chloro-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-6,6-dioxido-
7,8-dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl)acrylamide (24b)
Light yellow solid. 32.9% yield, m.p: 259.8-261.4 ºC. 1H NMR (400 MHz, DMSO-d6) δ 10.02 (s, 1H, -CO-NH-), 9.32 (s,

1H, -NH-), 8.41 (s, 1H, indole-H), 8.31 (d, J = 3.9 Hz, 1H, indole-H), 7.82(s, 1H, Ar-H), 7.61 (dt, J = 8.8, 4.2 Hz, 1H,
indole-H), 7.44–7.35 (m, 1H, Ar-H), 7.08 (tt, J = 9.3, 4.0 Hz, 1H, indole-H), 6.99 (d, J = 7.8 Hz, 1H, -C=CH2), 5.79 (d, J
= 8.8 Hz, 1H, -C=CH2), 4.59 (s, 2H, -SO2-CH2-), 3.91 (s, 3H, -OCH3), 3.83 (s, 3H, indole-CH3), 3.64 (d, J = 6.2 Hz, 3H,
-N-CH3), 3.41 (d, J = 6.8 Hz, 2H, -SO2-CH2-), 3.36 (s, 4H, -SO2-CH2-CH2-, -N-CH3-CH2-), 2.75 (d, J = 4.5 Hz, 6H,
N-(CH3)2), 2.61 (s, 2H, -CH2-). TOF MS ES+ (m/z): (M + H)+, calcd for C28H32FN7O2: 656.2222, found, 656.2238.
4.1.4.3.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-6,6-dioxido-
7,8-dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl)but-2-enamide (24c)
Light yellow solid. 35.6% yield, m.p: 257.9-259.4 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.83 (s, 1H, -CO-NH-), 8.89 (s,
1H, -NH-), 8.08 (d, J = 7.8 Hz, 1H, indole-H), 7.93 (d, J = 8.3 Hz, 1H, indole-H), 7.86 (d, J = 8.5 Hz, 1H, indole-H),
7.18 (s, 1H, Ar-H), 7.09(s, 1H, indole-H), 6.79 (s, 1H, Ar-H), 6.08 (s, 1H, -CO-CH=), 4.56 (s, 2H, -SO2-CH2-), 3.91 (s,
3H, -OCH3), 3.87 (s, 3H, indole-CH3), 3.77 (t, J = 6.9 Hz, 3H, -N-CH3), 3.33 (t, J = 7.1 Hz, 2H, -SO2-CH2-), 3.25–3.08
(m, 4H, -SO2-CH2-CH2-, -N-CH3-CH2-), 2.74 (s, 6H, N-(CH3)2), 2.63 (d, J = 13.1 Hz, 2H, -CH2-), 2.52 (s, 3H, =C(CH3)2,
2.28 (s, 3H, =C(CH3)2). 13C NMR (400 MHz, DMSO-d6) δ 166.31, 159.93, 159.28, 157.92, 155.12, 153.61, 147.07,
139.57, 135.04, 133.32, 128.21, 127.25, 126.84, 115.90, 115.25, 111.97, 111.80, 109.36, 107.89, 104.98, 97.79, 57.79,
56.15, 53.39, 52.15, 48.71, 45.74, 45.71, 38.75, 32.76, 29.61, 25.38, 20.76.
4.1.4.4.
N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-6,6-dioxido-7,8-
dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl)-3-methylbut-2-enamide (24d)
Light yellow solid. 33.9% yield, m.p: 258.1-260.7 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.86 (s, 1H, -CO-NH-), 8.48 (s,
1H, -NH-), 8.12 (s, 1H, indole-H), 7.83 (d, J = 9.9 Hz, 1H, indole-H), 7.59 (d, J = 5.8Hz, 1H, indole-H), 7.49–7.34 (m,
1H, Ar-H), 7.30 (d, J = 8.4 Hz, 1H, indole-H), 7.12 (t, J = 9.2 Hz, 1H, =CH-CH3), 7.03 (s, 1H, Ar-H), 6.83 (d, J = 14.9
Hz, 1H, -CO-CH=), 4.51 (s, 2H, -SO2-CH2-), 3.89(s, 3H, -OCH3), 3.83 (s, 3H, indole-CH3), 3.63 (s, 2H, -N-CH3-CH2-),
3.36–3.24 (m, 4H, -SO2-CH2-CH2-), 2.71 (s, 3H, -N-CH3), 2.60 (s, 2H, -CH2-), 2.46 (s, 6H, N-(CH3)2), 2.38 (s, 3H,
=CH-CH3). 13C NMR (400 MHz, DMSO-d6) δ 165.19, 159.93, 159.44, 157.92, 155.19, 147.07, 141.01, 139.57, 135.11,
133.39, 128.14, 127.25, 126.84, 121.26, 115.89, 111.88, 111.80, 109.36, 107.92, 104.98, 97.85, 57.71, 56.15, 53.42,
52.15, 48.79, 45.74, 45.65, 38.75, 32.76, 29.59, 16.79. TOF MS ES+ (m/z): (M + H)+, calcd for C28H32FN7O2: 650.2925,
found, 650.2931.
4.1.4.5.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-6,6-dioxido-7,8-
dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl)pent-2-enamide (24e)
Light yellow solid. 36.3% yield, m.p: 262.3.-263.5 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.49 (s, 1H, -CO-NH-), 8.39 (s,
1H, -NH-), 8.03 (s, 1H, indole-H), 7.79 (d, J = 10.3 Hz, 1H, indole-H), 7.50 (d, J = 5.6 Hz, 1H, indole-H), 7.42–7.32 (m,

1H, Ar-H), 7.26 (d, J = 8.5 Hz, 1H, indole-H), 7.05 (t, J = 8.7 Hz, 1H, =CH-CH2-), 6.96 (s, 1H, Ar-H), 6.78 (d, J = 15.3
Hz, 1H, -CO-CH=), 4.48 (s, 2H, -SO2-CH2-), 3.87 (s, 3H, -OCH3), 3.82 (s, 3H, indole-CH3), 3.58 (s, 2H, -N-CH3-CH2-),
3.30–3.19 (m, 4H, -SO2-CH2-CH2-), 2.67 (s, 3H, -N-CH3), 2.59 (s, 2H, -CH2-), 2.48 (s, 6H, N-(CH3)2), 2.23–2.14 (m, 2H,
-CH2-CH3), 1.04 (t, J = 7.6 Hz, 3H, -CH2-CH3). TOF MS ES+ (m/z): (M + H)+, calcd for C28H32FN7O2: 650.2925, found,
650.2928.
4.1.4.6.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-6,6-dioxido-7,8-
dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl)-4-methylpent-2-enamide (24f)
Light yellow solid. 35.4% yield, m.p: 263.8-265.2 ºC. 1H NMR (400 MHz, DMSO-d6) δ 10.12 (s, 1H, -CO-NH-), 9.42 (s,
1H, -NH-), 8.37 (s, 1H, indole-H), 8.02 (s, 1H, Ar-H), 7.79 (d, J = 10.6 Hz, 1H, indole-H), 7.50 (dd, J = 8.9, 4.6 Hz, 1H,
indole-H), 7.43–7.34 (m, 1H, indole-H), 7.07–7.01 (m, 1H, Ar-H), 6.95 (s, 1H, -CO-CH=), 6.73 (dd, J = 15.2, 6.2 Hz, 1H,
-CO-CH=CH), 4.48 (s, 2H, -SO2-CH2-), 3.87 (s, 3H, -OCH3), 3.83 (s, 3H, indole-CH3), 3.58 (t, J = 6.6 Hz, 3H, -N-CH3),
3.27 (d, J = 5.4 Hz, 6H, -N-CH3-CH2-, -SO2-CH2-CH2-), 2.73 (s, 6H, N-(CH3)2), 2.58 (s, 3H, -CH2-, -CH-(CH3)2), 1.04 (d,
J = 6.7 Hz, 6H, -CH-(CH3)2). 13C NMR (400 MHz, DMSO-d6) δ 164.97, 160.28, 159.93, 157.92, 155.17, 147.07, 146.74,
139.55, 135.04, 133.32, 128.04, 127.25, 126.84, 121.90, 115.90, 112.41, 111.82, 109.36, 107.77, 104.98, 97.81, 57.77,
56.15, 53.42, 52.15, 48.71, 45.74, 45.71, 38.75, 32.76, 31.23, 29.60, 21.52, 21.46. TOF MS ES+ (m/z): (M + H)+, calcd
for C28H32FN7O2: 664.3081, found, 664.3088.
4.1.4.6.
(Z)-N-(2-((2-(dimethylamino)ethyl)(methyl)amino)-5-((4-(5-fluoro-1-methyl-1H-indol-3-yl)-6,6-dioxido-7,8-
dihydro-5H-thiopyrano[4,3-d]pyrimidin-2-yl)amino)-4-methoxyphenyl)hex-2-enamide (24g)
Light yellow solid. 30.8% yield, m.p: 264.4-266.8 ºC. 1H NMR (400 MHz, DMSO-d6) δ 9.60 (s, 1H, -CO-NH-), 8.48 (s,
1H, -NH-), 8.37 (s, 1H, indole-H), 8.02 (s, 1H, Ar-H), 7.78 (d, J = 10.5 Hz, 1H, indole-H), 7.49 (dd, J = 8.9, 4.5 Hz, 1H,
indole-H), 7.43–7.32 (m, 1H, Ar-H), 7.04 (td, J = 9.0, 2.7 Hz, 1H, indole-H), 6.97 (s, 1H, -CO-CH=CH-), 6.72 (dt, J =
14.5, 6.9 Hz, 1H, -CO-CH=), 4.48 (s, 2H, -SO2-CH2-), 3.87 (s, 3H, -OCH3), 3.81 (s, 3H, indole-CH3), 3.58 (t, J = 6.7 Hz,
3H, -N-CH3), 3.28 (t, J = 6.7 Hz, 4H, N-CH3-CH2-, -SO2-CH2-), 3.11 (s, 2H, -SO2-CH2-CH2-), 2.62 (s, 2H, -CH2-), 2.54
(t, J = 10.4 Hz, 8H, N-(CH3)2, -CH2-CH2-CH3), 1.48–1.43 (m, 2H, -CH2-CH2-CH3), 0.90 (t, J = 7.3 Hz, 3H,
-CH2-CH2-CH3). TOF MS ES+ (m/z): (M + H)+, calcd for C28H32FN7O2: 664.3081, found, 664.3085.
4.2 Molecular docking study
The Sybyl 8.1 software packges (TRIPOS Associates Inc., USA) were applied to carry out the molecular docking
simulation. All the docking approaches used in the experiment can be found in the Tripos bookshelf. The crystal structure
of EGFR-T790M (PDB code: 3IKA) used in the docking were downloaded from the website http://www.rcsb.org/. The
protein preparation procedures implemented by using the Structure Preparation Tool in Sybyl, mainly including fixing the

residues, removing hetero-atoms and water molecules, adding hydrogen atoms and optimizing the whole structure of the
reconstruction protein, etc. The inhibitor 3 was first removed from the EGFR-T790M crystal structure. Then the target
compounds were put into the binding site. The surflex-dock algorithm was used to calculate the energy values between
the ligands and protein. Only the best-scoring complex was used for the further bingding site analysis. The Pymol 1.8x
(https://pymol.org). software was used to modfit and process the final docking results.
4.3. 3D-QSAR on Hela cell line
4.3.1. Data set
In this research, in order to further investigate the antiproliferative activity of target compounds against Hela cells, the in
vitro antiproliferative activity data of Hela cells were chose to construct the 3D-Quantitative Structure Activity
Relationship (3D-QSAR) models. Table 4 listed the structures of all target compounds and the IC50 values against Hela
cell. CoMFA and CoMSIA analysis were performed using IC50 values for all target compounds against Hela cells. Then,
the IC50 values were converted to pIC50 values according to the following formula:
1
pICꢀꢁ = Ig
ICꢀꢁ
4.3.2. Structure optimization
Construction of all target compounds, structural optimization, and 3D-QSAR modeling were all performed on Sybyl 8.1
(Tripos Associates Inc.) software. The Tripos force field and conjugate gradient method were used to minimize the
structure of all target compounds. The target charge was set to the Gasteiger-hücker charge. The convergence criterion
was 0.05 kcal/(mol·Å), and the maximum optimization frequency was 10000. Subsequently, Compound 23g with the
highest pIC50 value was chosen as the template molecule, and the skeleton N1-(7,8-dihydro-5H-thiopyrano[4,3-d]
pyrimidin-2-yl) benzene-1,3-diamine was set as alignment template to carry out molecular superposition.
4.3.3. CoMFA and CoMSIA studies
3D-QSAR analyses of target compounds were performed using the Sybyl 8.1 software default settings. Steric and
electrostatic fields energy (CoMFA mode) were calculated using a positively charged SP3 hybridization C atom with a
van der Waals force radius of 1.52 Å as the probe atom (default probe atom in Sybyl). The interaction energy values
between C atom and target compounds were calculated by calculating the interaction energy values of C atom and
compounds according to the Lennard-Jones and Coulomb potential energy functions with the threshold value of 30
KJ/mol.
According to the Lennard-Jones and Coulomb potential energy functions. The electrostatic field energy and the
steric field energy around the target compound were determined. The threshold was set to 30 kcal/mol. CoMSIA model