Lead generation of 1,2-dithiolanes as exon 19 and exon 21 mutant EGFR tyrosine kinase inhibitors

Full text of DOI:10.1016/j.bmcl.2019.04.029

Accepted Manuscript
Lead generation of 1,2-dithiolanes as exon 19 and exon 21 mutant EGFR tyro-
sine kinase inhibitors
Tarek S. Mansour, Vijay Potluri, Ranga R. Pallepati, Vishnu Basetti, Mallaiah
Keesara, Ashok G. Moghudula, Pranab Maiti
PII:
S0960-894X(19)30251-3
https://doi.org/10.1016/j.bmcl.2019.04.029
BMCL 26396
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
15 February 2019
15 April 2019
18 April 2019
Please cite this article as: Mansour, T.S., Potluri, V., Pallepati, R.R., Basetti, V., Keesara, M., Moghudula, A.G.,
Maiti, P., Lead generation of 1,2-dithiolanes as exon 19 and exon 21 mutant EGFR tyrosine kinase inhibitors,
Bioorganic & Medicinal Chemistry Letters (2019), doi: https://doi.org/10.1016/j.bmcl.2019.04.029
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.

Lead generation of 1,2-dithiolanes as exon 19 and exon 21 mutant EGFR tyrosine
kinase inhibitors
Tarek S. Mansour a, Vijay Potluri b, Ranga R. Pallepati b, Vishnu Basetti b, Mallaiah
Keesara b, Ashok G. Moghudula b, Pranab Maiti b
a Sabila Biosciences LLC, 5 Overlook Road, New City, New York 10956 U.S.A.
b Aurigene Discovery Technology Ltd. Bollaram Road, Miyapur, Hyderabad, 500 049,
Telangana India
Corresponding author: tsmansour@sabilabiosciences.com
Non-small cell lung cancer (NSCLC) is the most common type (80-85%) of lung cancer
sub-categorized into adenocarcinoma, squamous cell carcinoma and large cell
carcinoma. Unfortunately, the majority of NSCLC cases are diagnosed in patients
reaching localized advanced stage III or metastatic stage IV and thus require some form
of chemotherapy intervention. Over 160,000 deaths occurred in the USA in 2018 due to
respiratory system cancers including NSCLC.1 Initial therapy with the first-generation
tyrosine kinase inhibitors (TKIs) erlotinib, gefitinib and icotinib achieved remarkable
benefits targeting the endothelial growth factor receptor (EGFR) tyrosine kinase domain.
Mutations in EGFR which activate the RAS signaling pathway, occur to the extent of
85% in NSCLC due to exon 19 deletions (E746-A750del, ex19del) and exon 21 point
mutations such as (L858R). Gefitinib was first approved for primary NSCLC that
contains activating mutations EGFRL858R/ex19del. Second-generation TKIs such as
neratinib, canertinib, dacomitinib and afatinib were designed to covalently target a
conservative Cys797 residue in both wild type (WT) and mutant EGFR thus enhancing
their kinome selectivity and duration of response due to longer target engagement in the
presence of high intracellular concentrations of ATP. However, rapid emergence of
resistance to all clinically used first and second-generation TKIs in about 50% of
patients, limits their clinical benefit due to selection of a single gatekeeper mutation
T790M (exon 20) that increases affinity to ATP and restricts access to a hydrophobic
binding pocket. To overcome acquired resistance to EGFRT790M, third-generation
covalent TKIs,2-4 osimertinib and olmutinib were developed3 and a number of
experimental drugs such as nazartinib,5 naquotinib,6 avitinib,7 PF067477758 and others
are in various stages of development.9,10 Despite these impressive advances, multiple
resistance mechanisms to osimertinib and olmutinib have been reported most notably
dominated by the emergence of a C797S mutation in about 40% of clinical cases.2-4
Ironically, this mutation prevents the formation of a covalent bond between the inhibitors
and Cys797. The intense clinical and regulatory activities in the period of 2013-2018
resulted in the FDA approval of erlotinib, gefitinib, afatinib, osimertinib and olmutinib
covering different modalities to treat NSCLC patients. Current research efforts have
focused on further development of fourth-generation inhibitors with a profile of inhibiting,
activating mutations (L858R, exon19del), T790M and the C797S mutations
(EGFRL858R/T790M/C797S) and indeed, good progress has been made as in brigatinib11
1

(phase 2), the non-ATP allosteric inhibitor EA104512 and others,3,13,14 although it
remains to be seen whether these fourth-generation inhibitors will rapidly select for
resistance as well.
An alternative strategy to developing fourth-generation inhibitors is based on combining
EGFRL858R/exdel19 first-generation and EGFRT790M third-generation inhibitors. The
principle behind this strategy is supported by examining the allelic context of EGFRC797S
and EGFRT790M selected by a combination of osimertinib with erlotinib which determines
the sensitivity to the TKIs combination. Trans alleles (on the same clones) are sensitive
to first and third-generation TKIs combinations, whereas cis alleles (different clones) are
less effected.15,16 Structural analysis suggests that the third-generation TKIs restore the
cytoplasmic juxtamembrane segment structure to the conformation of EGFR in the
wild-type state when activated by its natural ligand EGF.17 This juxtamembrane
segment structure which links the kinase domain with the extracellular regions is likely
an important player since mutant and EGFRWT have similar kinase domains. We
therefore, considered a strategy to address the rapid resistance to third-generation
inhibitors by a de novo design of EGFRL858R/exdel19 inhibitors for combination therapy with
third-generation inhibitors. In this paper, we report on our lead identification efforts
which led to novel EGFRL858R/ex19del TKIs that carry 1,2-dithiolanes as a possible moiety
that might be targeted by the conservative Cys797 residue in an alternative fashion to
acrylamide containing inhibitors. Sparing EGFRWT activity remains a desirable property
of TKIs to limit skin rash side effects due to keratinocytes injury which eventually leads
to release of IL-31, a known pruritus-inducing cytokine.18
Quinazolines and 3-cyanoquinolines were chosen as core heterocycles due to their
precedents as EGFR inhibitors. Various N-4 headpiece moieties explored in this study
(Figure 1) were selected from those in clinical candidates
pelitinib/afatinib/canertinib/sapitinib/dacomitinib, lapatinib and CP-724,714. The key
intermediates: 6-aminoquinazolines and -3-cyanoquinolines (Figure 1, 1-8) were
prepared by a displacement reaction of the 4-chloro heterocycles with anilines as
previously described for 6.19
2

O
X
R
O
X
N
R
N
HN
N
1
; X=CH3, R=
; X=Cl, R=
HN
N
3
; X=CH3, R=
H2N
O
N
H2N
CN
2
4
; X=Cl, R=
F
F
Y
F
O
HN
Cl
5
6
; Y=F
7
; (R)
Cl
H2N
CN
HN
O
8
; Y=
; (S)
F
H2N
O
CN
O
N
N
O
O
Figure 1. Key Aniline Intermediates 1-8.
3-Cyanoquinolines 7 and 8 were prepared in 4 steps involving: 1) dealkylation of the 7-
methoxy heterocycle 9 with BBr3 in dichloromethane, 2) displacement of the 4-chloro
moiety in 10 with 3-chloro-4-((3-fluorobenzyl)oxy)aniline, 3) O-alkylation of 11 using
either (S) or (R)-(2,2-dimethyl-1,3-dioxolan-4-yl)methyl methanesulfonate with Cs2CO3
with N.N-dimethyl formamide , 4) hydrolysis with potassium hydroxide in methanol as
shown in Scheme 1.
3

Scheme 1. Synthesis Protocol for 7 and 8
O
F
O
Cl
N
O
O
Cl
N
HN
N
Cl
HN
O
CN
a, b
c
HN
HO
CN
HN
HO
CN
9
11
10
f, e, g
d, e, g
O
F
F
O
Cl
Cl
HN
HN
N
H2N
O
CN
H2N
O
CN
N
O
O
7
8
O
O
Reagents and conditions: (a) BBr3 DCM, 0 oC 0.5 h, (b) 100 oC, 5 h, 80% , (c) 3-chloro-4-
((3-fluorobenzyl)oxy) aniline, i-propanol, 150 oC 12 h 56%, (d) Cs2CO3 DMF, rt 0.5 ,
(e) (R)-(2,2-dimethyl-1,3-dioxolan-4-yl)methyl methanesulfonate 90 oC, 12 h, 12%,
(f) (S)-(2,2-dimethyl-1,3-dioxolan-4-yl)methyl methanesulfonate 90 oC, 12 h, 12%,
(g) KOH/MeOH 70 oC 6 h, 25%
Anilines 1-6 were coupled with various dithio derivatives using one of three methods A,
B or C to furnish target compounds 13 – 32 (Schemes 2, 3). A direct coupling protocol
of anilines 1, 2, 4, and 6 with a 1,2 dithiolane acyl chloride generated in situ from 2-(1,2-
dithiolan-3-yl)acetic acid 33 or lipoic acid 34 and oxalyl chloride in dichloromethane-
tetrahydrofuran solvent system was achieved in method A. In this fashion, target
compounds 16, 17, 19 and 22 were prepared.
In method B, a coupling agent such as 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide
hydrochloride (EDC) in N-methyl pyrrolidone was used to effect the reaction of anilines
3, 5, 6 with asparagusic acid 35 as well as 33 and 34. Method B proved to be useful for
the preparation of 13, 15, 21, 25 and 28. In an alternative approach depicted as method
C, 2-(bromomethyl)acrylic acid 36 was converted stepwise to 3-(acetylthio )-2-
((acetylthio)methyl)propanoic acid 38 by reaction with thioacetic acid via the
intermediacy of 37. Compound 38 was then converted to its chloroacyl derivative S,S'-
(2-(chlorocarbonyl)propane-1,3-diyl) diethanethioate 39 by reaction with oxalyl chloride.
Coupling of 39 with anilines 1, 2, 3, 5 and 6 produced the corresponding S,S'-
4

carbamoyl)propane-1,3-diyl)diethanethioates which were isolated and treated with
methanolic ammonia to provide compounds 14, 18, 20, 24, and 27. Conversion of 23
and 26 to 24 and 27 respectively was also achieved via hydrolysis of 23 and 26 with
sodium borohydride in dimethyl acetamide followed by acid quench to produce bis-
mercapto intermediates which were readily oxidized by dimethyl sulfoxide to
asparagusic analogs 24 and 27.
Scheme 2. Synthesis of TKIs by Methods A, B and C
Method A
O
OH
a, b
16, 17, 19, 22
n
S
S
; 33
n=1
34
n=4;
O
O
N
F
Cl
HN
HN
N
H
N
H
N
CN
N
S
O
S
S
O
S
O
N
F
O
O
17
16
19
F
Cl
Cl
HN
HN
H
H
N
N
CN
N
S
S
S
O
O
S
O
N
O
N
22
o
1, 2, 4, 6,
Reagents and conditions: (a) Oxalyl chloride, DCM-THF 0 C, 15 min, (b) Anilines
DIEA, DMF-THF 0 oC, 1 h, 8-13% 2 steps.
5

Method B
O
OH
or
O
a, b
S
S
n
13, 15, 21, 25, 28
S
S
OH
; 33
n=1
35
34
n=4;
F
O
O
N
F
Cl
Cl
HN
N
H
N
HN
S
S
HN
H
H
N
CN
N
CN
CN
S
S
O
S
O
O
S
O
N
N
21
13
15
F
F
O
Cl
Cl
HN
N
H
HN
N
N
CN
H
S
S
N
CN
O
O
S
O
S
O
25
28
3, 5, 6
Reagents and Conditions: (a) EDC, NMP (b) Anilines
rt 2 h, 17-37% 2 steps
6

Method C
O
O
COOH
Br
COOH
SAc
a, b, c
AcS
OH
AcS
Cl
SAc
d
e
SAc
36
38
37
39
f, g
h
23, 26
14, 18, 20
O
O
O
N
N
F
Cl
S
S
HN
S
S
S
S
HN
N
H
N
HN
H
N
H
N
CN
N
N
O
O
O
O
N
14
O
N
18
20
F
F
AcS
AcS
S
Cl
Cl
S
HN
N
HN
H
N
H
CN
N
CN
i
O
O
O
O
N
23
24
F
O
F
O
Cl
Cl
AcS
AcS
HN
N
S
S
H
HN
N
H
N
CN
i
N
CN
O
O
O
O
26
27
Reagents and conditions: (a) Na2CO3-H2O 47% (b) AcSH (c) HCl 45% (d) AcSH 45% (e) Oxalyl
o
1, 2, 3
38
chloride DCM-THF, DIEA 0 C, 1 h, (f) anilines
(g) NH3 in MeOH 2-4% 3 steps from
o
5 6
(h) anilines , 4% (i)
23 24
to
26 27
to
and
NaBH4 DMA (i) DMSO 120 C, 1 h, 12-17% 2 steps
7

Compounds 29-32 were prepared following the protocols in method C (Scheme 3) by
coupling of 7 and 8 with 38 to produce 39 and 40 respectively, followed by deprotecting
of acetonides with acid to diols 29 and 30. Oxidative hydrolysis to 1,2 dithiolanes 31 and
32 was achieved by methanolic ammonia in dichloromethane.
Scheme 3. Synthesis of 29-32
O
AcS
Cl
SAc
7
8
, a
, a
38
AcS
AcS
AcS
AcS
R
O
HN
N
R
O
HN
HN
O
CN
HN
O
CN
O
N
O
39
O
40
O
b
b
AcS
AcS
AcS
AcS
R
O
HN
R
O
HN
HN
O
CN
HN
CN
N
HO
HO
O
N
HO
29
30
HO
c
c
S
S
S
S
R
R
O
O
HN
HN
N
HN
CN
HN
O
CN
O
N
O
HO
HO
HO
31
32
F
HO
R =
Cl
Reagents and conditions: (a) Method C DIPEA-DCM, rt 2 h, 34-63% (b) HCl MeOH rt 30 min,
10% (c) NH3 MeOH rt 30 min, 75%
8

The SAR was assessed in two biochemical binding assays: rabbit reticulocyte lysate
and KINOMEscan™. Initially, inhibitors 13-32 were evaluated against EGFRWT and
EGFRL858R and Her2 kinases in the rabbit reticulocyte lysate assay (KinaseSeekerTM)
which contains the cellular components necessary for protein synthesis. This assay is a
homogeneous competition binding assay where the displacement of an active site
dependent probe by an inhibitor is measured by a change in luminescence signal
reported in IC50 values.20 Luminescence readout translates into a highly sensitive and
robust assay with low background and minimal interference from test compounds.
The activity of compounds in mutant EGFR was assessed in the KINOMEscan™ assay
which is a competition binding assay that quantitatively measures the ability of a
compound to compete with an immobilized, active-site directed ligand. The ability of the
test compound to compete with the immobilized ligand is measured via quantitative
PCR of the DNA tag and reported as inhibitor binding constants (Kd values) from
duplicate 11-point dose response curves.21
In the set of 3-cyanoquinolines (13-16) lacking any C7 substituent, 13 and 14 displayed
weaker activity against EGFRWT and EGFRL858R than 15 and 16 suggesting that the 3-
chloro-4-((3-fluorobenzyl)oxy)anilino headpiece is preferred over the 3-methyl-4-((6-
methylpyridin-3-yl)oxy)anilino moiety of 13 and 14. This finding was consistently
observed in the 7-methoxyquinazolines 17-20 where 19 and 20 emerged 8-60 fold more
potent than 17 and 18 against EGFRL858R .Preference for inhibiting EGFRL858R by
inhibitors containing the 3-chloro-4-((3-fluorobenzyl)oxy)anilino over 3-chloro-4-
fluoroanilino headpieces was established in several pairs of lipoic acid conjugates 21
and 22, asparagusic acid conjugates 24 and 27 and, 2-(1,2-dithiolan-3-yl)acetic acid
conjugates 25 and 28. From this set of inhibitors, we deduced that asparagusic and 2-
(1,2-dithiolan-3-yl)acetic acid conjugates have similar activity against EGFRL858R and
both conjugates displayed good selectivity over EGFRWT. Acyclic intermediates 23 and
26 were about 2-fold less active against EGFRL858R than target compounds 24 and 27
respectively. The 7-dihydroxypropyl analogs 31 and 32 maintained the increased
inhibition of EGFRL858R with good selectivity over EGFRWT (Table 1). In this pair, small
preference for the S stereochemistry was noted. Compounds 15, 21, 23-25 were not
evaluated against EGFRWT due to their modest activity against EGFRL858R
.
9

Table 1. SAR in Rabbit Reticulocyte Assay (IC50 Values in nM)*
Entry
1
Compound
13
EGFRWT
8630 ± 850
>10,000
NT
164 ± 15
253 ± 20
658 ± 219
25 ± 2
76 ± 6
NT
730 ± 65
NT
EGFRL858R
>10,000
>10,000
199 ± 19
36 ± 12
48 ± 8
413 ± 68
5.8 ±1
6.5 ± 0.8
420 ± 5
33 ± 4
494 ± 30
219 ± 47
300 ± 40
26.8 ± 4.2
18 ± 3
Method
B
C
B
A
A
C
A
C
B
A
C
C
B
C
C
B
C
C
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
31
NT
NT
340 ± 67
210 ± 20
120 ± 19
125 ± 12
256 ± 21
10.9 ± 1.9
9 ± 2
32
20 ± 7
*Ten point dose response curve, each value is an average of triplicate experiments
It was deemed of relevance to benchmark the inhibitors against a select set of clinical
and marketed compounds including mutant selective as well as first-generation
inhibitors. Since historical data in the rabbit reticulocyte assay were limited for these
inhibitors, the IC50 values (in nM) were determined in this assay format (Supplementary
Figure 1). Inhibitor 28 compared well to this set and was considered worthy of further
evaluation.
In order to determine the activity of lead candidates in mutant EGFR enzymes other
than EGFRL858R, the KINOMEscan™ assay was chosen because it had the ability to test
all relevant mutants of EGFR. Thus, we first established the correlation between the
rabbit reticulocyte and KINOMEscan™ assays by comparing the IC50 and Kd values for
three most potent inhibitors 22, 26 and 28 against EGFRL858R. The data (Supplementary
Table 1) showed excellent correlation between the two assay formats thus justifying the
use of this assay to determine the activity in mutant EGFR.
Next, the activity of lead candidate 28 was determined against several activating EGFR
mutations including the highly frequent deletion mutation of EGFRex19del. The Kd values
were in the range of 4.8-48 nM against ten EGFR variants including the EGFRL861Q
,
10

EGFRG719C and EGFRex19del mutants (Supplementary Table 2). Inhibitor 28 displayed
good activity (48 nM) against the EGFRex19del mutant, being about 2-4 fold better in
comparison to the asparagusic acid analog 27 (110 nM) which contains the isomeric
1,2-dithiolanyl moiety and its precursor 26 (180 nM) respectively.
To evaluate kinase selectivity, Her2 the closely related ErbB family member was
chosen. In this case all EGFRL858R inhibitors tested displayed good selectivity (> 25 fold)
over Her2 in the same rabbit reticulocyte lysate assay format with 17 being the least
selective (Table 2). Moreover, 28 displayed even greater selectivity (Supplementary
Table 3) over several TEC family kinases (TEC, BMX, BTK, and ITK).
Table 2. IC50* Values (nM) against Her2
Entry
Compound
EGFRWT
8630 ± 850
164 ± 15
253 ± 20
658 ± 219
210 ± 20
120 ± 19
125 ± 12
EGFRL858R
>10,000
36 ± 12
48 ± 8
413 ± 68
18 ± 3
Her2
1
2
3
4
5
6
7
13
16
17
18
27
28
31
>10,000
730 ± 150
93 ± 19
590 ± 155
730 ± 150
160 ± 20
167 ± 20
10.9 ± 1.9
9 ± 2
*Ten point dose response curve, each value is an average of triplicate experiments
Prior to this letter, the kinome activity of 1,2 dithiolane carboxamides was rather
limited.22 Melchiorre et al,23 proposed an intriguing multitarget-directed drug discovery
strategy by combining structural features of oxidative stress and EGFR molecular
recognition elements represented by 41 (Supplementary Figure 2) which contains a 6-
substituted 1,2-dithiolanyl carboxamide with a quinazoline core. Based on reversibility
data in human epidermoid carcinoma A431 cells which overexpress EGFR and
molecular modeling studies, the investigators concluded that 41 irreversibly block EGFR
by a thiol-disulfide interchange reaction involving Cys797 of EGFR and the disulfide
moiety of 41. We have utilized scanKINETICTM assay to evaluate the
association/dissociation kinetics of two inhibitors 31 and 30. In this assay, the inhibitor
binding kinetics are determined by measuring and comparing the Kd values under four
sets of dose-response: 1) arm A: full equilibrating; 2) arm B: equilibrate/3-fold
dilution/equilibrate; 3) arm C: partial equilibration and 4) arm D: predilution/full
equilibration. The Kd values for 31 against EGFRL858R under these four conditions was
determined to be 0.7, 8.4, 1.9 and 16 nM. This suggests that 31 kinetics are consistent
with a reversible inhibitor with rapid dissociation kinetics (Kd arm B/ Kd arm A =12) and
relatively slow association kinetics (Kd arm C/ Kd arm A =2.7). Furthermore, compound
30 which is an open chain analog of 31 had Kd values against EGFRL858R in the
scanKINETICTM assay as follows: 2.1, 58, 5.9 and 60 nM in arms A-D respectively.
These data suggest that relative to 31, compound 30 has even faster dissociation
11

kinetics but similar slow association kinetics (Supplementary Table 4). Therefore, on the
basis of these findings, it was concluded that 30 and 31 do not bind EGFRL858R in an
irreversible manner.24 The apparent difference between our results and those of
Melchiorre et al regarding the involvement of a possible thiol-disulfide interchange
reaction may be due to the assay format (biochemical versus cellular), choice of
dithiolane substrate with 4 versus 3-substituted dithiolanes, lack of C7 substituent or the
nature of the kinase (EGFRL858R versus EGFRWT).
Inhibitor 28 has acceptable CYP isoforms 1A2, 3A4, 2B6, 2C8, 2C9, 2C19, 2D6 and
metabolic profiles in S9, liver microsomes and hepatocytes (Supplementary Tables 5
and 6). In the NCI-60 panel, 28 as well as 17 demonstrated good cellular activity relative
to gefitinib against three NSCLC cell lines and two leukemia lines with GI50 values in the
range of 0.11-1.07 µM (Table 3). The desirable high LC50 values obtained resulted in
excellent therapeutic index (2->227) in this set of cell lines.
Table 3. Cellular Activity in µM of 17, 28 and Gefitinib in Select Cells in the NCI-60
Panela
17
LC50
28
LC50
Gefitinib
b
c
b
c
e
Cell Line
NSCLC
GI50
TId
GI50
TId
GI50
A549
0.76
1.07 > 50
0.11 NT
NT
0.54
> 46.7 0.60
0.47
3.4
> 25
1.95
6.2
> 41.5
4.1
7.94
0.08
6.31
NCI-H322M
NCI-H460
Leukemia
CCRF-CEM
MOLT-4
0.22 > 50
0.16 > 2.6
> 227
16.2
0.15
0.16
> 25
> 25
> 166
> 156
5.01
3.98
(a) Five dose response curve, each value is an average of 2 experiments; (b) GI50 is the
concentration of the drug causing 50% growth inhibition (c) LC50 is the cytotoxic
concentration due to 50% reduction in measured protein (d) LC50/GI50 (e) data from the
DTP-NCI database https://dtp.cancer.gov/services/nci60data/meangraph/gi50/jpg/-
4.0/715055
In conclusion, we have outlined a new synthetic strategy to EGFR mutant selective
inhibitors that led to the discovery of a new series of 1,2 dithiolane TKIs with potent
activity against exon 19 deletion and exon 21 mutant EGFR. These inhibitors are
selective over EGFRWT, Her2 and TEC family kinases in the rabbit reticulocyte lysate
assay. Two inhibitors 17 and 28 demonstrated good cellular activity and high
therapeutic index in certain NSCLC and leukemia cell lines. Binding to EGFRL858R is
shown to be reversible as determined by the scanKINETICTM assay. The lead candidate
28 has good metabolic and excellent CYP isoform profile. Further optimization of lead
candidate 28 shall be reported in due course.24
12

Acknowledgements: We thank the NCI DTP program (Bethesda, MD, USA) for cellular
data, Reena Zutshi (Luceome Biotechnology, Tucson, AZ, USA)) and Jeremy Hunt
(Eurofins, San Diego, CA, USA)) for all biochemical data, R Mamatha and S Kumar for
CYP and metabolic stability data. We also thank Paul C. Anderson and Mahdi B. Fawzi
for useful discussions on this research. This paper is dedicated with appreciation to Elie
Abushanab (retired Professor) University of Rhode Island, RI, USA.
This research did not receive any specific grant from funding agencies in the public,
commercial, or not-for-profit sectors
Declaration of interest: None
References and notes
1. Siegel RI, Miller KD, Jemal A. Cancer stats 2017. CA Cancer. J Clin. 2017;67:7-30.
2. Song, Z, Ge Y, Wang C, Huang S, Shu X, Liu K, Zhou Y, Ma X. Challenges and
perspectives on the development of small-molecule EGFR inhibitors against T790M-
mediated resistance in non-small-cell lung cancer. J Med Chem. 2016;59:6580–
6594.
3. Chen L, Fu W, Zheng L, Liu Z, Liang G. Recent progress of small-molecule
epidermal growth factor receptor (EGFR) inhibitors against C797S in non-small-cell
lung cancer. J Med Chem. 2018;61:4290–4300.
4. Cheng H, Planken S. Precedence and promise of covalent inhibitors of EGFR and
KRAS for patients with non-small-cell lung cancer. ACS Med Chem Lett.
2018:9:861-863.
5. Lelais G, Epple R, Marsilje TH, Long YO, McNeill M, Chen B, Lu W, Anumolu J,
Badiger S, Bursulaya B, DiDonato M, Fong R, Juarez J, Li J, Manuia M, Mason
DE, Gordon P, Groessl P, Johnson K, Jia Y, Kasibhatla S, Li C, Isbell J, Spraggon
G, Bender S, Michellys PY. Discovery of (R,E)-N-(7-chloro-1-(1-[4-
(dimethylamino)but-2-enoyl]azepan-3-yl)-1H-benzo[d]imidazol-2-yl)-2-
methylisonicotinamide (EGF816), a novel, potent, and WT sparing covalent inhibitor
of oncogenic (L858R, ex19del) and resistant (T790M) EGFR mutants for the
treatment of EGFR mutant non-small-cell lung cancers. J Med Chem. 2016;59:6671
–6689.
6. Wang S, Cang S, Lin D. Third-generation inhibitors targeting EGFR T790M mutation
in advanced non-small cell lung cancer. J Hematol Oncol. 2016;9:34.
7. Xu X. Parallel phase 1 clinical trials in the US and in China accelerating the test of
avitinib in lung cancer as a novel inhibitor selectively targeting mutated EGFR and
overcoming T790M-induced resistance. Clin J Cancer. 2015;34:285-287.
13

8. Planken S, Behenna DC, Nair SK, Johnson TO, Nagata A, Almaden C, Bailey S,
Baillard TE, Bernier L, Cheng H, Cho-Schultz S, Dalvie D, Deal JG, Dinh DC,
Edwards MP, Ferre RA, Gajiwatla KS, Hemkens M, Kania RS, Kath JC, Matthews J,
Murray BW, Niessen S, Orr STM, Pairsih M, Sach NW, Shen H, Shi M, Solowieij J,
Tran K, Tseng E, Vicini P, Wang Y, Weinrich SL, Zhou R, Zientek, M, Liu L, Luo Y,
Xin S, Zhang C, Lafontaine J. Discovery of N-((3R,4R)-4-fluoro-1-(6-((3-methoxy-1-
methyl-1H-pyrazol-4-yl)amino)-9-methyl-9H-purin-2-yl)pyrrolidine-3-yl)acrylamide
(PF-06747775) through structure-based drug design: a high affinity irreversible
inhibitor targeting oncogenic EGFR mutants with selectivity over wild-type EGFR. J
Med Chem. 2017;60:3002–3019.
9. Hao Zhang H, Wang J, Shen Y, Wang HY, Duan WM, Zhao HY, Hei, YY, Xin M,
Cao YX, Zhang SQ. Discovery of 2,4,6-trisubstituted pyrido[3,4-d]pyrimidine
derivatives as new EGFR-TKIs. Eur J Med Chem. 2018;148:221-237.
10.Chan BK, Hanan EJ, Bowman KK, Bryan MC, Burdick D, Chan E, Chen Y, Clausen
S, Dela Vega T, Dotson J, Eigenbro C, Elliot RL, Heald RA, Jackson PS, Knight JD,
La H, Lainchbury MD, Malek S, Purkey HE, Schaefer G, Schmidt S, Seward EM,
Sideris S, Shao L, Wang S, Yeap SK, Yen I, Yu C, Heffron TP, Discovery of a
noncovalent, mutant-selective epidermal growth factor receptor inhibitor. J Med
Chem. 2016;59:9080–9093.
11.Uchibori K, Inasse N, Araki M, Kamada M, Sato S, Okuno Y, Fujita N, Katayama R.
Brigatinib combined with anti-EGFR antibody overcomes osimertinib resistance in
EGFR-mutated non-small cell lung cancer. Nat Commun. 2017:8:14768.
12.Jia Y, Yun CH, Park E, Ercan D, Manuai M, Juarez J, Xu C, Rhee K, Chen T, Zhang
H, Palakurthi S, Jang J, Lelais G, DiDonato M, Bursulaya B, Michellys, PY, Epple, R,
Marsilije, TH, McNeill M, Lu W, Harris J, Bender S, Wong KK, Janne PA, Eck MJ.
Overcoming EGFR(T790M) and EGFR(C797S) resistance with mutant-selective
allosteric inhibitors. Nature 2016;534:129-132.
13.Gűnther M, Lategahn J, Juchum M, Döring E, Keul M, Engel J, Tumbrink HL, Rauh
D, Laufer S. Trisubstituted pyridinylimidazoles as potent inhibitors of the clinically
resistant L858R/T790M/C797S EGFR mutant: targeting of both hydrophobic regions
and the phosphate binding site. J Med Chem. 2017;60: 5613–5637.
14.Lu X, Zhang T, Zhu S-J, Xun Q, Tong L, Hu X, Li Y, Chan S, Su Y, Sun Y, Chen Y,
Ding J, Yun C-H, Xie H, Ding K. Discovery of JND3229 as a new EGFRC797S mutant
inhibitor with in vivo mono drug efficacy. ACS Med.Chem.Lett. 2018;9:1123-1127.
15.Niederst MJ, Hu H, Mulvey HE, Lockerman EL, Garcia AR, Piotrowska Z, Sequist
LV, Engelman JA. The allelic context of the C797S mutation acquired upon
treatment with third-generation EGFR inhibitors impacts sensitivity to subsequent
treatment strategies. Clin Cancer Res. 2015;21:3924-3933.
14

16.Wang Z, Yang, JL, Huang J, Ye JY, Zhang XC, Tu HY, Han-Zhang H, Wu YL. Lung
adenocarcinoma harboring EGFR T790M and in trans C797S responds to
combination therapy of first- and third-generation EGFR TKIs and shifts allelic
configuration at resistance. J Thoracic Oncol. 2017;12:1723-1727.
17.Lowder MA, Doener AE, Schepartz A. Structural differences between wild-type and
double mutant EGFR modulated by third-generation kinase inhibitors. J Am Chem
Soc. 2015;137:6456–6459.
18.Gangemi S, Franchina T, Mincuillo PL, Profita M, Zanghi M, David A, Kennez I,
Adamo, V. IL-33/IL-31 axis: a new pathological mechanism for EGFR tyrosine
kinase inhibitors-associated skin toxicity. J Cell Biochem. 2013;114: 2673-2676.
19.Tsou HR, Overbeek-Klempers EG, Hallet WA, Reich MF, Floyd MB, Johnson BD,
Michalak RS, Nilakantan R, Discafani C, Golas J, Rabindran SK, Shen R, Shi X,
Wang YF, Upeslacis J, Wissner, A. Optimization of 6,7-disubstitiuted-4-
(arylamino)quinolone-3-carbonitriles as orally active, irreversible inhibitors of human
epidermal growth factor receptor-2 kinase activity. J Med Chem. 2005;48: 1107-
1131.
20.Jester BW, Cox KJ, Gaj A, Shomin CD, Porter JR, Ghosh I. A coiled-coil enabled
split-luciferase three-hybrid system: applied toward profiling inhibitors of protein
kinases. J Am Chem Soc. 2010;132:11727-11735.
21.Karaman MW, Herrgard S, Triber DK, Gallant P, Atteridge CE, Campbell, BT, Chan
KW, Ciceri P, Davis MI, Edeen PL, Faraoni R, Floyd M, Hunt JP, Lockhart DJ,
Milanov ZV, Morrison MJ, Pallares G, Patel HK, Pritchard S, Woodicka LM, Zarrinkar
PA. Quantitative analysis of kinase inhibitor selectivity. Nat Biotechnol.
2008;26:127–132.
22.Zhang WJ, Wei H’ Hagen T’ Frei, B. α-Lipoic acid attenuates LPS-induced
inflammatory responses by activating the phosphoinositide 3-kinase/Akt signaling
pathway. Proc Natl Acad Sci. U.S.A. 2007;104:4077-4082
23.Antonello A, Tarozzi A, Morroni F, Cavalli A, Rosini M, Hrelia P, Bolognesi ML,
Melchiorre C. Multi-target directed drug design strategy: a novel molecule designed
to block epidermal growth factor receptor (EGFR) and to exert proapoptotic effects. J
Med Chem. 2006;49:6642-6645.
24.Beno BR, Yeung, K-S, Bartberger, MD, Pennington, LD, Meanwell, NA. A survey of
the role of noncovalent sulfur interactions in drug design. J Med Chem.
2015;58:4383-4438.
15

25.For role of 1,2-dithiolane derivatives in DNA damage, see Nejad MI, Yang D, Shen
B, Gates KS. Oxidative activation of leinamycin E1 triggers alkylation of guanine
residues in double-stranded DNA. Chem Commun. 2018;54:256-259.
F
O
O
N
S
S
Cl
HN
N
S
S
H
N
HN
N
H
CN
N
CN
O
O
O
14
28
IC50 (nM)
EGFRL858R >10,000
IC50 (nM)
Kd (nM)
EGFRL858R 11
EGFRex19del 48
EGFRL861Q 4.8
EGFRWT
120
160
EGFRHer2
EGFR719C
5.7
FIGURE LEGENDS
Figure 1. Key Aniline Intermediates 1-8
16

O
X
R
O
X
N
R
N
HN
N
1
; X=CH3, R=
; X=Cl, R=
HN
N
3
; X=CH3, R=
H2N
O
N
H2N
CN
2
4
; X=Cl, R=
F
F
Y
F
O
HN
Cl
5
6
; Y=F
7
; (R)
Cl
H2N
CN
HN
O
8
; Y=
; (S)
F
H2N
O
CN
O
N
N
O
O
Figure 1. Key Aniline Intermediates 1-8.
SCHEMES LEGENDS
Scheme 1. Synthesis Protocol for 7 and 8.
Scheme 2. Synthesis of TKIs by Methods A, B and C
Scheme 3. Synthesis of 29-32
17

Scheme 1. Synthesis Protocol for 7 and 8
O
F
O
Cl
N
O
O
Cl
N
HN
N
Cl
HN
O
CN
a, b
c
HN
HO
CN
HN
HO
CN
9
11
10
f, e, g
d, e, g
O
F
F
O
Cl
Cl
HN
HN
N
H2N
O
CN
H2N
O
CN
N
O
O
7
8
O
O
Reagents and conditions: (a) BBr3 DCM, 0 oC 0.5 h, (b) 100 oC, 5 h, 80% , (c) 3-chloro-4-
((3-fluorobenzyl)oxy) aniline, i-propanol, 150 oC 12 h 56%, (d) Cs2CO3 DMF, rt 0.5 ,
(e) (R)-(2,2-dimethyl-1,3-dioxolan-4-yl)methyl methanesulfonate 90 oC, 12 h, 12%,
(f) (S)-(2,2-dimethyl-1,3-dioxolan-4-yl)methyl methanesulfonate 90 oC, 12 h, 12%,
(g) KOH/MeOH 70 oC 6 h, 25%
Scheme 2. Synthesis of TKIs by Methods A, B and C
18

Method A
O
OH
a, b
16, 17, 19, 22
n
S
S
; 33
n=1
34
n=4;
O
O
N
F
Cl
HN
HN
N
H
N
H
N
CN
N
S
O
S
S
O
S
O
N
F
O
O
17
16
19
F
Cl
Cl
HN
HN
H
H
N
N
CN
N
S
S
S
O
O
S
O
N
O
N
22
o
1, 2, 4, 6,
Reagents and conditions: (a) Oxalyl chloride, DCM-THF 0 C, 15 min, (b) Anilines
DIEA, DMF-THF 0 oC, 1 h, 8-13% 2 steps.
19

Method B
O
OH
or
O
a, b
S
S
n
13, 15, 21, 25, 28
S
S
OH
; 33
n=1
35
34
n=4;
F
O
O
N
F
Cl
Cl
HN
N
H
N
HN
S
S
HN
H
H
N
CN
N
CN
CN
S
S
O
S
O
O
S
O
N
N
21
13
15
F
F
O
Cl
Cl
HN
N
H
HN
N
N
CN
H
S
S
N
CN
O
O
S
O
S
O
25
28
3, 5, 6
Reagents and Conditions: (a) EDC, NMP (b) Anilines
rt 2 h, 17-37% 2 steps
20

Method C
O
O
COOH
Br
COOH
SAc
a, b, c
AcS
OH
AcS
Cl
SAc
d
e
SAc
36
38
37
39
f, g
h
23, 26
14, 18, 20
O
O
O
N
N
F
Cl
S
S
HN
S
S
S
S
HN
N
H
N
HN
H
N
H
N
CN
N
N
O
O
O
O
N
14
O
N
18
20
F
F
AcS
AcS
S
Cl
Cl
S
HN
N
HN
H
N
H
CN
N
CN
i
O
O
O
O
N
23
24
F
O
F
O
Cl
Cl
AcS
AcS
HN
N
S
S
H
HN
N
H
N
CN
i
N
CN
O
O
O
O
26
27
Reagents and conditions: (a) Na2CO3-H2O 47% (b) AcSH (c) HCl 45% (d) AcSH 45% (e) Oxalyl
o
1, 2, 3
38
chloride DCM-THF, DIEA 0 C, 1 h, (f) anilines
(g) NH3 in MeOH 2-4% 3 steps from
o
5 6
(h) anilines , 4% (i)
23 24
to
26 27
to
and
NaBH4 DMA (i) DMSO 120 C, 1 h, 12-17% 2 steps
21

Scheme 3. Synthesis of 29-32
O
AcS
Cl
SAc
7
8
, a
, a
38
AcS
AcS
AcS
AcS
R
O
HN
N
R
O
HN
HN
O
CN
HN
O
CN
O
N
O
39
O
40
O
b
b
AcS
AcS
AcS
AcS
R
O
HN
R
O
HN
HN
O
CN
HN
CN
N
HO
HO
O
N
HO
29
30
HO
c
c
S
S
S
S
R
R
O
O
HN
HN
N
HN
CN
HN
O
CN
O
N
O
HO
HO
HO
31
32
F
HO
R =
Cl
Reagents and conditions: (a) Method C DIPEA-DCM, rt 2 h, 34-63% (b) HCl MeOH rt 30 min,
10% (c) NH3 MeOH rt 30 min, 75%
TABLE LEGENDS
22

Table 1. SAR in Rabbit Reticulocyte Assay (IC50 Values in nM)
Table 2. IC50 Values (nM) Against Her2 in the Rabbit Reticulocyte Lysate Assay
Table 3. Cellular Activity in µM of 17 and 28 in the NCI-60 Panel
Table 1. SAR in Rabbit Reticulocyte Assay (IC50 Values in nM)*
Entry
1
Compound
13
EGFRWT
8630 ± 850
>10,000
NT
164 ± 15
253 ± 20
658 ± 219
25 ± 2
76 ± 6
NT
730 ± 65
NT
EGFRL858R
>10,000
>10,000
199 ± 19
36 ± 12
48 ± 8
413 ± 68
5.8 ±1
6.5 ± 0.8
420 ± 5
33 ± 4
494 ± 30
219 ± 47
300 ± 40
26.8 ± 4.2
18 ± 3
Method
B
C
B
A
A
C
A
C
B
A
C
C
B
C
C
B
C
C
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
31
NT
NT
340 ± 67
210 ± 20
120 ± 19
125 ± 12
256 ± 21
10.9 ± 1.9
9 ± 2
32
20 ± 7
*Ten point dose response curve, each value is an average of triplicate experiments
Table 2. IC50* Values (nM) against Her2
Entry
Compound
EGFRWT
8630 ± 850
164 ± 15
253 ± 20
658 ± 219
210 ± 20
120 ± 19
125 ± 12
EGFRL858R
>10,000
36 ± 12
48 ± 8
413 ± 68
18 ± 3
Her2
1
2
3
4
5
6
7
13
16
17
18
27
28
31
>10,000
730 ± 150
93 ± 19
590 ± 155
730 ± 150
160 ± 20
167 ± 20
10.9 ± 1.9
9 ± 2
*Ten point dose response curve, each value is an average of triplicate experiments
23

Table 3. Cellular Activity in µM of 17, 28 and Gefitinib in Select Cells in the NCI-60
Panela
17
LC50
28
LC50
Gefitinib
b
c
b
c
e
Cell Line
NSCLC
GI50
TId
GI50
TId
GI50
A549
0.76
1.07 > 50
0.11 NT
NT
0.54
> 46.7 0.60
0.47
3.4
> 25
1.95
6.2
> 41.5
4.1
7.94
0.08
6.31
NCI-H322M
NCI-H460
Leukemia
CCRF-CEM
MOLT-4
0.22 > 50
0.16 > 2.6
> 227
16.2
0.15
0.16
> 25
> 25
> 166
> 156
5.01
3.98
(a) Five dose response curve, each value is an average of 2 experiments; (b) GI50 is the
concentration of the drug causing 50% growth inhibition (c) LC50 is the cytotoxic
concentration due to 50% reduction in measured protein (d) LC50/GI50 (e) data from the
DTP-NCI database https://dtp.cancer.gov/services/nci60data/meangraph/gi50/jpg/-
4.0/715055
24