Dual nicotinamide phosphoribosyltransferase and epidermal growth factor receptor inhibitors for the treatment of cancer

Article abstract of DOI:10.1016/j.ejmech.2020.113022

Multitarget drugs have emerged as a promising treatment modality in modern anticancer therapy. Taking advantage of the synergy of NAMPT and EGFR inhibition, we have developed the first compounds that serve as dual inhibitors of NAMPT and EGFR. On the basis of CHS828 and erlotinib, a series of hybrid molecules were successfully designed and synthesized by merging of the pharmacophores. Among the compounds that were synthesized, compound 28 showed good NAMPT and EGFR inhibition, and excellent in vitro anti-proliferative activity. Compound 28, which is a new chemotype devoid of a Michael receptor, strongly inhibited the proliferation of several cancer cell lines, including H1975 non-small cell lung cancer cells harboring the EGFR^L858R/T790M mutation. More importantly, it imparted significant in vivo antitumor efficacy in a human NSCLC (H1975) xenograft nude mouse model. This study provides promising leads for the development of novel antitumor agents and valuable pharmacological probes for the assessment of dual inhibition in NAMPT and EGFR pathway with a single inhibitor.

Full text of DOI:10.1016/j.ejmech.2020.113022

Journal Pre-proof
Dual Nicotinamide Phosphoribosyltransferase and Epidermal Growth Factor Receptor
Inhibitors for the Treatment of Cancer
Wanheng Zhang, Kuojun Zhang, Yiwu Yao, Yunyao Liu, Yong Ni, Chenzhong Liao,
Zhengchao Tu, Yatao Qiu, Dexiang Wang, Dong Chen, Lei Qiang, Zheng Li, Sheng
Jiang
PII:
S0223-5234(20)30994-6
DOI:
https://doi.org/10.1016/j.ejmech.2020.113022
EJMECH 113022
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 18 July 2020
Revised Date: 17 October 2020
Accepted Date: 11 November 2020
Please cite this article as: W. Zhang, K. Zhang, Y. Yao, Y. Liu, Y. Ni, C. Liao, Z. Tu, Y. Qiu, D. Wang,
D. Chen, L. Qiang, Z. Li, S. Jiang, Dual Nicotinamide Phosphoribosyltransferase and Epidermal Growth
Factor Receptor Inhibitors for the Treatment of Cancer, European Journal of Medicinal Chemistry,
https://doi.org/10.1016/j.ejmech.2020.113022.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
© 2020 Published by Elsevier Masson SAS.

Dual Nicotinamide Phosphoribosyltransferase and Epidermal Growth Factor
Receptor Inhibitors for the Treatment of Cancer
Wanheng Zhanga,#, Kuojun Zhanga,#, Yiwu Yaoa,#, Yunyao Liua,#, Yong Nia,
Chenzhong Liaob, Zhengchao Tuc, Yatao Qiua, Dexiang Wanga, Dong Chena, Lei
Qiang a, Zheng Lid,*, Sheng Jianga,**
aState Key Laboratory of Natural Medicines, Jiangsu Key Laboratory of Drug Design
and Optimization and Department of Medicinal Chemistry, China Pharmaceutical
University, Nanjing, 210009, China
bSchool of Biological and Medical Engineering, Hefei University of Technology,
Hefei, 230009, China
cGuangzhou Institutes of Biomedicine and Health, Chinese Academy of Sciences,
Guangzhou 510530, China
dCenter for Bioenergetics, Houston Methodist Research Institute, 6670 Bertner,
Houston, Texas 77030, United States

Abstract: Multitarget drugs have emerged as a promising treatment modality in
modern anticancer therapy. Taking advantage of the synergy of NAMPT and EGFR
inhibition, we have developed the first compounds that serve as dual inhibitors of
NAMPT and EGFR. On the basis of CHS828 and erlotinib, a series of hybrid
molecules were successfully designed and synthesized by merging of the
pharmacophores. Among the compounds that were synthesized, compound 28 showed
good NAMPT and EGFR inhibition, and excellent in vitro anti-proliferative activity.
Compound 28, which is a new chemotype devoid of a Michael receptor, strongly
inhibited the proliferation of several cancer cell lines, including H1975 non-small cell
lung cancer cells harboring the EGFRL858R/T790M mutation. More importantly, it
imparted significant in vivo antitumor efficacy in a human NSCLC (H1975) xenograft
nude mouse model. This study provides promising leads for the development of novel
antitumor agents and valuable pharmacological probes for the assessment of dual
inhibition in NAMPT and EGFR pathway with a single inhibitor.
Keywords
: NAMPT, EGFR, multitarget drugs, dual inhibitor, antiproliferative
activity, antitumor efficacy

1. Introduction
Cancer is a multifactorial disease associated with dysregulation of multiple genes and
cellular signaling pathways. Traditionally targeted therapies that are designed to act
specifically on a single oncogenic target often have limited benefits to cancer patients.
Rationally, simultaneous modulation of a network of cancer-related targets could be
an effective therapeutic approach, and would lead to extensive research and
application of specific drug combinations. However, the employment of drug
cocktails has some shortcomings and weaknesses, including complex and
unpredictable pharmacokinetic properties, potential drug-drug interactions, and poor
patient compliance.1, 2 In an attempt to overcome these disadvantages, a single
molecule with multiple targets has captured considerable attention and emerged as an
important treatment modality in modern cancer therapy.3, 4 Thus, multi-targeted drugs
can simultaneously and often synergistically target different signaling pathways
implicated in the disease resulting in additional effects and better efficacy. In addition,
such drugs can have the advantage of concurrent pharmacokinetic profiles, minimized
drug-drug interactions and reduced toxicity.
Metabolic alterations are a prototypical hallmark of tumor development which
has become an active research area for cancer treatment.5 Among these targets,
nicotinamide adenine dinucleotide (NAD) metabolism has attracted much attention
due to its indispensable role in supporting cancer cell proliferation.6 NAD is a critical
component involved in multiple cellular energy-generating and signaling transduction
processes and ultimately determining cellular fate. In mammalian cells, NAD can be

synthesized mainly through four pathways: (i) de novo biosynthesis employing
tryptophan as starting material, (ii) the primary salvage pathway starting from
nicotinamide (NAM), and two alternative salvage pathways using (iii) nicotinic acid
(NA) and (iv) nicotinamide riboside (NR) as precursors, respectively.7 Since the
NAM-dependent process enables the efficient recycling of NAM by NAD-consuming
enzymes such as sirtuins and poly-ADP-ribose polymerases (PARPs), this pathway is
the major source for the replenishment and maintenance of NAD.8 Nicotinamide
phosphoribosyltransferase (NAMPT) is the rate-limiting enzyme responsible for the
conversation of NAM into nicotinamide mononucleotide (NMN) and the subsequent
synthesis of NAD.9, 10 Because of the uncontrolled and rapid proliferation, cancer
cells have much higher energy demand and elevated expression or activity of PARPs
and sirtuin than normal cells. In addition, cancer cells have an apparent increase of
reactive oxygen species (ROS). NAD and its reduced form NADH contribute to
maintaining redox balance of the tumor microenvironment to prevent cells from the
oxidative injury.11, 12 Cancer cells that consume more NAD and are more dependent
on NAD-mediated events are more susceptible to NAMPT inhibition than normal
cells.13 Normal cells can use an alternative nicotinic acid phosphoribosyltransferase
(NAPRT)-mediated pathway for NAD synthesis from nicotinic acid (NA), thereby
protecting themselves from NAMPT inhibition,14, 15 while most cancer cells lack
NAPRT activity.16, 17 NAMPT is therefore considered to be an attractive target for
selective cancer therapy18-20 and to date, a number of NAMPT inhibitors have been
reported. Among these, FK866 (1)21, 22 and CHS828 (2)23, 24 have advanced into phase

II clinical trials (Figure 1A).25 However, their clinical applications have been
restricted due to their dose-limiting thrombocytopenia and gastrointestinal toxicity
and other undesirable pharmacokinetic properties.26, 27 Therefore, there is an urgent
need for innovative NAMPT inhibitors with reduced side effects.
EGFR is a member of the ErbB family of receptor tyrosine kinases (RTK) that is
a transmembrane protein containing a ligand-binding domain on the external surface
of the cytomembrane, a hydrophobic transmembrane region and a tyrosine domain
extending into the cytoplasm28, 29. Notably, EGFR dysregulation through
overexpressions or activating mutations play an important role in tumor development
and progression, and has been confirmed to be one of the most valuable targets for the
31
treatment of cancer, especially non-small-cell lung cancer (NSCLC).30,
First-generation inhibitors such as gefitinib (3) and erlotinib (4) (Fig. 1) are currently
used as the front-line standard therapy in treating NSCLC patients harboring EGFR
activating mutation such as L858R mutation and the exon-19 deletion.32-35 Despite
initial encouraging results, most treated patients eventually relapse and develop
secondary or acquired resistance after continuous treatment.36-38 The most prevalent
acquired resistance mechanism is the EGFR mutation T790 M, accounting for
50~60% of the cases39. The second-generation irreversible EGFR inhibitors, such as
afatinib (5)40-42 and dacomitinib (6)43, 44, and third-generation selective irreversible
inhibitors，such as WZ4002 (7),45 osimertinib (8)46, 47 and olmutinib (9)45, 48, have
been shown to effectively overcome the resistance induced by a T790M mutation
through forming a covalent bond between an inherent electrophilic Michael acceptor

and a conserved Cys797 near the ATP binding domain of EGFR.49 However,
additional resistance rapidly occurs, with one of the potential mechanisms being the
appearance of tertiary C797S mutation50-52. Therefore, continuous research efforts are
urgently needed to discover new drug candidates in order to improve patient
outcomes.
We envisioned that the discovery of new chemical entities that are devoid of a
Michael acceptor, but active against EGFR T790M mutation should be an effective
approach to relieve drug resistance induced by C797S mutation. On the other hand,
drug combinations and multitargeted agents have been proposed to combat the
acquired resistance and enhance the antitumor efficacy of RTK inhibitors. It has been
reported that the growth of EGFR-gene-mutated NSCLC, including activating,
T790M and C797S mutations, requires amounts of intracellular ATP to activate
EGFR-driven signaling transduction.53 As described above, NAMPT is fundamentally
important in energy generation and maintenance of ATP levels in mammalian cells.
Consequently, we postulated that NAMPT inhibitors might be able to synergistically
enhance the inhibitory effect of EGFR inhibitors. A single molecule that
simultaneously or concurrently can inhibit NAMPT and EGFR enzyme activity would
possess a promising anticancer profile. This molecular hybridization strategy that
combines two synergistic pharmacophores directly or via a linker in order to act on
different targets has emerged as a paradigm in drug design. Dual-targeting inhibitors
that act by merging the pharmacophore of NAMPT inhibitors into active agents
targeting histone deacetylase (HDAC) have been evaluated in cancer treatment.54, 55

Herein, we report the first-in-class NAMPT/EGFR bifunctional inhibitors that are
structurally lack of a Michael acceptor designed and synthesized on the basis of the
aforementioned evidence and multi-targeted drug theory. Particularly,
(Z)-2-cyano-1-(6-((4-((3-cyanophenyl)-amino)-7-methoxyquinazolin-6-yl)oxy)pentyl
)-3-(pyridin-4-yl)guanidine (28) exhibits concurrent inhibition against NAMPT and
EGFR with excellent in vitro antiproliferative activity and in vivo antitumor efficacy.

Cl
H
H
H
N
N
N
N
O
N
O
N
N
O
N
1
(FK866 / WK175 / APO866)
2
(CHS828 / GMX1778)
F
O
O
N
O
O
N
O
HN
Cl
N
O
HN
N
O
N
4
(erlotinib)
3
(gefitinib)
O
F
O
O
N
O
N
O
N
HN
N
Cl
H
N
H
N
N
N
Cl
N
H
N
N
N
HN
O
O
N
N
H
O
F
Cl
5
(WZ4002)
6 (afatinib)
7 (dacomitinib)
O
O
H
N
N
N
N
O
N
N
H
N
S
N
N
N
N
NH
N
H
O
9
(olmutinib)
8
(osimertinib)
Fig. 1. Representative NAMPT and EGFR inhibitors.
2. Results and discussion
2.1. Rational design of NAMPT/EGFR bifunctional inhibitors
As shown in Fig. 2C, a series of dual NAMPT/EGFR inhibitors were designed
utilizing a molecular hybridization strategy. Compared with direct bridging of two
inhibitors with a chain linker, a fused pharmacophore is more desirable owing to the
likelihood of a lower molecular weight. The NAMPT inhibitor (2) and the EGFR
inhibitor (4) were chosen as the templates for this design. Most NAMPT inhibitors are
generally linear and characterized by a widely accepted pharmacophore model
containing a cap group mimicking NAM, a connecting unit, a linker and a
hydrophobic tail group (Fig. 2A) 18, 25. The cap group combined with the connecting
unit constitutes the core, which is important for NAMPT inhibition. The linker should
8

have an appropriate length and geometry, allowing the tail group to protrude toward
the solvent-exposed surface and interact with the hydrophobic amino acids in the rim
of the enzyme. However, the tail group can be tolerated since the hydrophobic cleft of
the enzyme is large enough to accommodate extraordinarily variable structures.25, 56
Accordingly, for the NAMPT inhibitor (2), the core pyridinylcyanoguanidine motif is
indispensable for enzyme inhibitory activity, while the hydrophobic group of 2 can be
replaced with a fragment of erlotinib (4) to occupy the deep hydrophobic pocket of
NAMPT. We explored the erlotinib substructure and found by analysis of the X-ray
co-crystal structure of EGFR complexed with erlotinib that it can be modified without
sacrificing its EGFR inhibitory activity.
Quinazoline is a privileged skeleton for EGFR inhibitors. From the co-crystal
structure of EGFR with erlotinib, we saw that the quinazoline motif occupies the
adenine region of the ATP binding pocket and forms two important hydrogen bonds at
N1 and N3, and the phenylamino group fits well into the hydrophobic pocket.57, 58
Therefore, these two moieties cannot be replaced. The ethereal side chains at C-6 and
C-7 of the quinazoline ring protrude outward the receptor (Fig. 2B).57, 58 From the
perspective of structure-activity relationships (SARs), modification at C-6 and C-7
sites of the quinazoline ring is tolerable and the pharmacophore of 2 introduced to
these positions was postulated to retain EGFR/HER2 binding affinity. Meanwhile, the
phenylaminoquinazoline backbone of the EGFR inhibitor should fit well into the
hydrophobic pocket of the NAMPT enzyme, and the attached side chain with the
terminal pyridinylcyanoguanidine was predicted to mimic NAM as required to
9

suppress the NAMPT activity. Thus, we merged pyridinylcyanoguanidine into the
quinazoline pharmacophore of EGFR inhibitor in an attempt to produce, in a single
compound, the dual inhibitory activities against NMAPT and EGFR (Fig. 2C).
Fig. 2. Schematic illustration of the rationale underlying the design of dual NAMPT/EGFR inhibitors.
(A) Pharmacophore model of NAMPT inhibitor 2; (B) Pharmacophore model of EGFR inhibitor 4; (C)
The fusion of pharmacophore of NAMPT and EGFR inhibitors.
2.2. Chemistry
The synthetic routes of analogues with various chain lengths and pyridine
substitutions are depicted in Schemes 1，2 and 3. The synthesis began with the
protection of commercially available starting materials 10, 11 with a
t-butyloxycarbonyl (Boc) group. Treatment of 13-15 with triphenylphosphine and
carbon tetrabromide afforded halohydrocarbons 14 and 15 via an Appel reaction.
Aminolysis of N-cyanoimido-S,S-dimethyl-dithiocarbonate with primary amines 16
10

and 17 in the presence of NaH provided intermediates 18 and 19, respectively, in
good yields.
The quinazoline moiety of erlotinib was prepared from the commercially
available compound (20), which was first reacted with thionyl chloride using DMF as
catalyst to give the 4-chloroquinazoline derivative (21). Nucleophilic aromatic
substitution of 21 with 3-aminophenylacetylene generated intermediate 22, which
subsequently underwent hydrolysis with ammonium hydroxide to afford an alcohol
(23).
The target compounds (26-29) were synthesized as shown in Scheme 2. In an
alkaline condition, intermediate 23 was installed in the Boc-protected
halohydrocarbon amine linkers 14 and 15 to furnish 24 and 25. Deprotection of
compounds 24 and 25 using trifluoroacetic acid, followed by reaction with
N'-cyano-N-(pyridine-4-yl)- carbamimidothioate (18) or methyl N'-cyano-N-
(pyridinyl)-carbamimidothioate (19) gave the final products 26-29.
As shown in Scheme 3, compound 23 was coupled with commercially available
7-bromoheptanenitrile to provide an intermediate which has seven methylene spacers,
and which was further reacted with LiAlH4 to afford the corresponding amine.
Subsequently, treatment of the amine with methyl N'-cyano-N-(pyridine-4-yl)-
carbamimidothioate (18) or methyl N'-cyano-N-(pyridine-3-yl)carbamimidothioate
(19) gave the final compounds 30 and 31, respectively, with good yields.
11

Scheme 1. Synthesis of key intermediates 14, 15, 18, 19 and 23a
H
H
b
a
H2N
OH
n
N
Br
n
N
OH
n
Boc
Boc
12: n = 6
13: n = 5
10: n = 6
11
14: n = 6
15
: n = 5
: n = 5
H
c
N
R
S
R
NH2
N
N
16: R = 4-pyridyl
17: R = 3-pyridyl
18: R = 4-pyridyl
19: R = 3-pyridyl
Cl
OH
HN
N
AcO
AcO
AcO
O
N
N
d
e
N
O
N
O
N
20
21
22
HN
HO
O
f
N
N
23
a Reagents and conditions: (a) (Boc)2O, DIPEA, DCM, 0 °C - rt, 3 h, 82%; (b) CBr4, PPh3, DCM, 0° C - rt, 16 h,
75% ; (c) NaH, N-cyanoimido-S,S-dimethyl-dithiocarbonate, DMF, 0 °C - rt, 69%; (d) SOCl2, DMF, reflux,
overnight, 89%; (e) 3-aminophenylacetylene, isopropanol, reflux, 83%; (f) ammonium hydroxide, MeOH, rt, 90%.
Scheme 2. Synthesis of target compounds 26-29a
Boc
HN
HN
N
R
HN
N
H
O
n
HN
HN
N
O
n
N
N
HO
O
a
N
N
O
b
O
N
N
24
: n = 6
25: n = 5
26: R = 4-pyridyl, n = 6 27: R = 3-pyridyl, n = 6
28 29
23
: R = 4-pyridyl, n = 5
: R = 3-pyridyl, n = 5
a Reagents and conditions: (a) 14-15, K2CO3, DMF, 0 °C - rt, 68-85%; (b) (i) TFA, DCM, 0 °C - rt, (ii) 18 or 19,
DMAP, TEA, pyridine, 50 °C, 42-50% in two steps.
12

Scheme 3. Synthesis of target compounds 30-31a
a Reagents and conditions: (a) (i) 7-bromoheptanenitrile, K2CO3, DMF, 0 °C - rt; (ii) LiAlH4, THF, 0 °C - rt; (iii)
18 or 19, DMAP, TEA, pyridine, 50 °C, 34-41% in three steps.
2.3. In vitro NAMPT and EGFR inhibition assay
Initially, we evaluated the inhibitory activities of these compounds against NAMPT,
EGFR, EGFRT790M, EGFRL861Q and EGFRL858R. The results, shown in Table 1,
showed potency in the nanomolar range against EGFR, EGFRL861Q, EGFRL858R
,
EGFRT790M and NAMPT. Retaining the carbon chain length of six methylenes in
CHS828 (2) in the linking section, compounds 26 and 27 showed excellent inhibitory
activity against EGFR, EGFRL861Q and EGFRL858R, but somewhat reduced efficacy
against NAMPT and EGFRT790M. To gain good inhibitory effects against NAMPT and
EGFRT790M, we attempted to modify the carbon chain linker between the quinazoline
C-6 oxygen and the nitrogen atom of pyridinylcyanoguanidine. Lengthening the
distance to a seven- methylene spacer led to an NAMPT and EGFRT790M inhibition
similar to that of the compound with six methylenes, while shortening the distance to
five methylenes yielded a significantly increased NMAPT inhibition (28, IC50 = 41.2
nM; 29, IC50 = 74 nM), and slightly increased EGFRT790M inhibition (28, IC50 = 153
nM). In addition, the previous SAR studies of CHS828 (2) highlighted the importance
of the pyridine ring.59 Therefore, we further explored the position of the substituent in
13

Table 1. In vitro NAMPT and EGFR inhibitory activities of the target compounds
NAMPT
GlideScorec
EGFR
EGFRT790M
IC50 (nM)
NT
EGFRL861Q
IC50 (nM)
NT
EGFRL858R
IC50 (nM)
NT
Compounds
IC50 (nM)
20.7±3.11
NT
IC50 (nM)
NTb
GlideScore
NT
1(FK866)
NT
5 (WZ4002)
NT
4.59±0.46
0.35±0.03
0.32±0.031
0.23±0.014
0.52±0.0
0.473±0.0
0.538±0.0
NT
1.98±0.23
323.0±43.5
326.0±39.3
153.0±16.2
228.0±15.8
581.0±34.9
448.0±36.7
2.86±0.23
0.36±0.038
0.38±0.029
0.26±0.019
0.46±0.05
0.55±0.049
0.49±0.039
4.64±0.33
0.32±0.028
0.38±0.031
0.26±0.015
0.41±0.023
0.46±0.039
0.48±0.061
131.0±8.55
133.0±14.3
41.2±3.8
74.0±5.7
154.0±13.8
152.0±15.9
-5.7
-10.58
-9.74
26
27
28
29
30
31
-6.47
-8.83
-6.89
-5.98
-4.98
-10.42
-9.6
-10.64
-9.14
a IC50 values represent the mean of at least three independent experiments. bNT = not tested. cGlideScore was exported from Glide 6.7, lower scores mean higher binding affinity.
14

pyridyl ring. There was slight difference between 3-pyridyl and 4-pyridyl for both
targets with six- or seven- methylene linker (Table 1). The 4-pyridyl rather than the
3-pyridyl substitution resulted in potency increase with a spacer of five methylenes.
Of all the synthesized compounds, the 4-pyridyl compound (28) exhibited the highest
potency against EGFR, EGFRL861Q, EGFRL858R, EGFRT790M and NAMPT.
2.4. In vitro antiproliferation assay
Having obtained the potent in vitro NAMPT and EGFR inhibitory activities, we
further evaluated the antiproliferative activities of these compounds against Huh-7
(human liver cancer), MCF-7 (human breast cancer) and K562 (human myelogenous
leukemia) utilizing the CCK8 (Cell Counting Kit-8) assay. As shown in Table 2, most
of these compounds exhibited excellent inhibitory activity against human cancer cell
lines. Among all the tested tumor cell lines, the human breast cancer cell line MCF-7
was the most sensitive to all the compounds, delivering IC50 values in the nanomolar
range. Despite slight differences in enzyme inhibition between 4-pyridyl and
3-pyridyl analogues, the 4-pyridyl compounds had considerably more antiproliferative
activity than the 3-pyridyl compounds in various tumor cell lines. Compound 26 was
most active against human myelogenous leukemia k562 cell line with an IC50 value of
7.47 nM, and compound 28 was the most potent against the two solid tumor cell lines
(Huh-7 and MCF-7).
Table 2. Antiproliferative activities of compounds 26-31.
15

GI50 (nM)a
K562
Compd
Huh-7
MCF-7
26
27
28
29
1.87±0.17
8.02±1.10
1.64±0.13
62.9±6.92
7.47±0.68
24.6±2.58
38.0±3.11
175.0±13.69
0.15±0.01
0.16±0.007
0.10±0.006
0.98±0.07
30
31
1.81±0.15
9.94±1.21
7.41±0.96
50.9±1.02
42.1±4.37
2.76±0.43
0.21±0.018
0.63±0.04
4.85±0.61
Taxol
a GI50 values represent the mean of at least three independent experiments.
2.5. Molecular Docking
In an investigation of the binding modes of the target compounds in NAMPT and
EGFR, the docking program of Glide 6.7 standard precision model showed that
compound 28 could be docked into the active sites of NAMPT and wild-type EGFR.
The proposed binding modes of compound 28 in NMAPT and wild-type EGFR are
shown in Fig. 3. Compound 28 is accommodated very well in the active site of
NAMPT. The cap group of 28, i.e., the (E)-2-cyano-1-(pyridin-4-yl)guanidine moiety,
has identical interactions with this protein as the corresponding moiety of CHS828 (2).
The pyridine ring exhibits π-π offset stacking with Phe193, and could also have a
hydrogen bond interaction bridged by a water molecule with the protein. The guanidyl
group forms double hydrogen bonds with Asp219, while the cyano group forms a
16

hydrogen bond with Ser275. The tail group of compound 28, an
N-(3-ethynylphenyl)quinazolin-4-amine, has hydrophobic interactions with several
surrounding residues, including Tyr188, Val 242, Ile309, Ala379 (Fig. 3), similar to
the interactions between CHS828 and NAMPT.
The docking model demonstrated that the bulky tail group, the erlotinib segment
of compound 28 has identical interactions with that of erlotinib in the active site of
EGFR as described above (Fig. 3B). The head group, the (E)-2-cyano-1-(pyridin-
4-yl)guanidine moiety, could extend into a hydrophobic subpocket in the cleft of the
active site through carbon linkers. In this way, the pyridine moiety may have
additional interactions with surrounding residues such as Phe699, Leu834, and
Leu838. In addition, the polar 2-cyanoguanidine moiety forms four hydrogen bonds
with Lys721, Asp831 and Asn818 (Fig. 3B). In general, the docking modes are
consistent with the in vitro activities.
17

Fig. 3. (A) Overlap of the binding modes of CHS828 (green) and compound 28 (yellow) in NAMPT
(PDB code 4O12). (B) Overlap of the binding modes of erlotinib (green) and compound 28 (yellow)
EGFR (PDB code 4HJO).
2.6. In vivo antitumor efficacy
As mentioned above, compound 28 exhibited excellent inhibitory activity toward
both NAMPT and EGFR, and more importantly, possessed most potent
antiproliferative activity against Huh7 and MCF-7, cell lines which overexpress
wild-type EGFR. We further assessed its antiproliferative activity against the H1937
cell line harboring an EGFRL858R/T790M mutation. Compound 28 showed excellent
antiproliferative activity against the H1937 cell line with an IC50 value of 78.3 nM.
The in vivo antitumor efficacy of compound 28 in the H1937 xenograft nude mouse
model was evaluated. An H1975 lung cancer xenograft nude mouse model was
established to evaluate the in vivo antitumor potency of compound 28, and gefitinib, a
first-generation EGFR inhibitor was used as a negative control. The BALB/c nude
mice were inoculated in the right flank with suspensions of H1975 cells (3×106), and
when the mean tumor size reached approximately 150 mm3, compound 28 was
administered via i.v. injection 5 days a week for 3 weeks at a dose of 60 mg/kg. The
18

results of tumor growth inhibition in different groups at different time points after
treatment are shown in Fig. 4. Evidently, gefitinib was ineffective against H1937
xenograft (Fig. 4) and led to a significant loss of body weight (Fig. S1 in Supporting
Information, SI). Compound 28, however, significantly inhibited the tumor growth at
a dose of 60 mg/kg and delivered tumor volume reductions of 75.6%, relative to the
control group (Fig. 4A). Meanwhile, the tumor weights of mice treated with 28 at 60
mg/kg were reduced by 75% when compared with the control (Fig. 4B). In addition,
the toxicity of 28 was assessed by monitoring the body weight and survival of the
mice. Compound 28 was well tolerated and no significant body weight change was
observed in mice treated with 28 during the treatment period (Fig. S1).
Fig. 4. Compound 28 inhibited the growth of implanted H1975 xenograft in nude mice. H1975
tumor-bearing mice were treated with vehicle and 28 (60 mg/kg) with gefitinib (60 mg/kg) as the
negative control. (A) Changes in tumor volume of H1975 tumor-bearing mice during 3 weeks
treatment. (B) Tumor weight after 3 weeks treatment. Data are expressed as the mean ± standard
deviation. Statistical difference was determined by Student’s t-test. *P<0.05, compared with the vehicle
group. #P<0.05, compared with the gefitinib-treated groups.
3. Conclusion
Utilizing a pharmacophore merging strategy, we have successfully designed and
synthesized a series of first-in-class dual NAMPT/EGFR inhibitors. Optimization of
the linker between the quinazoline of erlotinib and the pyridinylcyanoguanidine of
19

CHS828 led to the identification of compound 28, a highly effective lead compound
with balanced NAMPT and EGFR inhibition. Interestingly, compound 28 exhibited
excellent antiproliferative activity in a panel of tumor cell lines, especially breast
cancer MCF-7 cell lines at low nanomolar concentrations (IC50 = 0.116 nM). Further,
it was tested against H1975 xenograft tumors that possess EGFRL858R/T790M mutation
in a nude mouse model. The results showed that compound 28 significantly retarded
the tumor growth in vivo. Third-generation EGFR inhibitors are known to have a
Michael receptor warhead component, which can form a covalent bond with the
active thiol of Cys797 in the EGFR ATP binding domain.50 However, mutant EGFR
with C797S lacks Cys797 and cannot form a covalent bond with third-generation
EGFR inhibitors, which would provide a key mechanism of resistance.50 Structurally,
28 does not have a Michael receptor but possesses significant efficacy against
EGFRL858R/T790M mutation, suggesting its potential to overcome the drug resistance
resulting from the T790M mutation.
In conclusion, we have for the first time prepared a dual inhibitor that
simultaneously reacts with both NAMPT and EGFR. This is a promising approach to
cancer chemotherapy, and further evaluation and optimization of the
erlotinib/CHS828 hybrids is now in progress.
4. Experimental section
4.1. General
Reagents and solvents from commercial sources were used without further
purification. The progress of all reactions was monitored by TLC using
20

EtOAc/n-hexane as solvent system, and spots were visualized by irradiation with UV
light (254 nm) or staining with phosphomolybdic acid. Flash chromatography was
1
13
performed using silica gel (300−400 mesh). H NMR and C NMR spectra were
recorded on a Bruker Avance ARX-400 or a Bruker Avance ARX-500. Chemical
shifts δ are reported in ppm, and multiplicity of signals are denoted as: s = singlet, d =
doublet, t = triplet and m = multiplet. The low resolution ESIMS was recorded on an
Agilent 1200 HPLC-MSD mass spectrometer and the high resolution on an Applied
Biosystems Q-STAR Elite ESI-LC-MS/MS mass spectrometer. Anhydrous
dichloromethane (DCM) and N,N-dimethylformamide (DMF) were freshly distilled
from calcium hydride. All other solvents were reagent grade. All moisture sensitive
reactions were carried out in flame dried flask under argon atmosphere. The purity of
the final compounds was determined by Agilent 1260 series HPLC system using the
following conditions: C-18 column (DicKma, 4.6 mm × 250 mm) with the solvent
system (elution conditions: mobile phase A consisting of MeOH; mobile phase B
consisting of water containing 0.1% ammonia), with monitoring between 190 and 800
nm. A flow rate of 1.0 mL/min was used. The retention time was reported as tR (min).
The purity of final compounds is > 95%.
4.2. Experimental procedures
4.2.1.
2-Cyano-1-(5-((4-((3-ethynylphenyl)amino)-7-methoxyquinazolin-6-yl)
oxy)-pentyl)-3-(pyridin-4-yl)guanidine (28)
Trifluoroacetic acid (38 ml) was added dropwise at 0 °C to a solution of 25 (19 g,
21

38.5 mmol) in anhydrous CH2Cl2 (150 ml). The reaction mixture was allowed to
warm to rt and then stirred until the materials were consumed. The reaction mixture
was concentrated in vacuo and then saturated NaHCO3 was added. The aqueous layer
was extracted twice with CH2Cl2 (100 ml) and the organic layer washed with brine
and dried over anhydrous Na2SO4. The solvent was removed under vacuum to give
the amine as a yellow solid which was used in the next step.
The above intermediate amine was dissolved in pyridine (100 ml), and compound
18 (7.4 g, 38.5 mmol), 4-dimethylaminopyridine (0.46g, 3.8 mmol), triethylamine
(10.7 ml, 77.0 mmol) were added. After stirring at 50 °C overnight, the reaction
mixture was cooled to rt. The pyridine was evaporated under reduced pressure and
then H2O (100 ml) was added. The aqueous layer was extracted twice with EtOAc
(100 ml), and the organic layer washed with brine and dried over anhydrous Na2SO4,
filtered and concentrated under reduced pressure. The resulting mixture was purified
by silica column chromatography and eluted with CH2Cl2/MeOH (40:1) to give 28
1
(8.4 g, 42% for two steps) as a faint yellow solid. H NMR (400 MHz, DMSO-d6) δ
9.47 (s, 1H), 9.38 (s, 1H), 8.49 (s, 1H), 8.37 (s, 2H), 7.98 (s, 1H), 7.89 (d, J = 8.2 Hz,
2H), 7.82 (s, 1H), 7.40 (t, J = 7.9 Hz, 1H), 7.21 (d, J = 9.2 Hz, 4H), 4.18 (s, 1H), 4.15
(d, J = 6.9 Hz, 2H), 3.93 (s, 3H), 3.33 (d, J = 8.2 Hz, 2H), 1.9-1.77 (m, 2H), 1.74-1.58
13
(m, 2H), 1.53-1.52 (m, 2H). C NMR (125 MHz, DMSO-d6) δ 156.5, 154.9, 153.1,
150.6, 150.0, 148.8, 147.4, 146.2, 140.2, 129.3, 126.8, 125.2, 123.0, 122.1, 116.9,
115.1, 109.3, 107.7, 103.0, 83.9, 80.9, 69.1, 56.3, 42.1, 29.0, 28.7, 23.4. HRMS (ESI)
calcd m/z for C29H28N8O2 [(M + H)+] 521.2408, found 521.2405.
22

4.2.2.
2-Cyano-1-(6-((4-((3-ethynylphenyl)amino)-7-methoxyquinazolin-6-yl)oxy)hexyl)-3-(p
yridin-4-yl)guanidine (26).
The procedure used was the same as described above for the synthesis of
compound 28. Compound 26 was obtained as a faint yellow solid (83 mg, 45% for
1
two steps). H NMR (400 MHz, DMSO-d6) δ 9.56 (s, 1H), 8.50 (s, 1H), 8.38 (d, J =
5.6 Hz, 2H), 7.99 (s, 2H), 7.89 (d, J = 8.3 Hz, 1H), 7.84 (s, 1H), 7.40 (t, J = 7.9 Hz,
1H), 7.31 – 6.99 (m, 4H), 4.19 (s, 1H), 4.15 (t, J = 6.5 Hz, 2H), 3.93 (s, 3H),
3.35-3.30 (m, 2H), 1.91-1.76 (m, 2H), 1.64-1.54 (m, 2H), 1.55-1.33 (m, 4H). 13C
NMR (125 MHz, DMSO-d6) δ 156.6, 155.0, 152.9, 149.6, 148.8, 146.9, 140.1, 129.3,
126.8, 125.3, 123.1, 122.1, 116.8, 114.9, 109.3, 107.4, 103.0, 83.9, 80.9, 69.2, 56.3,
42.2, 29.0, 29.0, 26.4, 25.7. HRMS (ESI) calcd m/z for C30H30N8O2 [(M + H)+]
535.2565, found 535.2567. HPLC analysis: 99.2% in purity.
4.2.3.
2-Cyano-1-(6-((4-((3-ethynylphenyl)amino)-7-methoxyquinazolin-6-yl)oxy)hexyl)-3-(p
yridin-3-yl)guanidine (27)
The procedure was the same as was used for the synthesis of compound 28.
1
Compound 27 was obtained as a yellow solid (55 mg, 50% for two steps). H NMR
(500 MHz, DMSO-d6) δ 9.50 (s, 1H), 9.20 (s, 1H), 8.49 (s, 1H), 8.46 (d, J = 2.6 Hz,
1H), 8.31 (d, J = 4.7 Hz, 1H), 7.99 (t, J = 1.9 Hz, 1H), 7.89 (dd, J = 8.2, 2.2 Hz, 1H),
7.83 (s, 1H), 7.66 (d, J = 8.2 Hz, 1H), 7.54 (t, J = 5.7 Hz, 1H), 7.39 (t, J = 7.9 Hz, 1H),
7.35 (dd, J = 8.3, 4.7 Hz, 1H), 7.20 (d, J = 11.4 Hz, 2H), 4.18 (s, 1H), 4.14 (t, J = 6.6
23

Hz, 2H), 3.93 (s, 3H), 3.26 (q, J = 6.7 Hz, 2H), 1.84 (d, J = 12.1 Hz, 2H), 1.62-1.53
13
(m, 2H), 1.53-1.45 (m, 2H), 1.41-1.39 (m, 2H). C NMR (125 MHz, DMSO-d6) δ
158.4, 156.5, 154.9, 153.1, 148.8, 147.3, 145.7, 145.2, 140.2, 135.1, 131.2, 129.3,
126.7, 125.2, 124.0, 123.0, 122.1, 117.4, 109.3, 107.7, 102.9, 83.9, 80.9, 69.2, 56.3,
42.0, 29.2, 29.0, 26.4, 25.7. HRMS (ESI) calcd m/z for C30H30N8O2 [(M + H)+]
535.2565, found 535.2566. HPLC analysis: 95.6% in purity.
4.2.4.
2-Cyano-1-(5-((4-((3-ethynylphenyl)amino)-7-methoxyquinazolin-6-yl)oxy)pentyl)-3-(
pyridin-3-yl)guanidine (29)
The procedure was the same as described above for the synthesis of compound 28.
Compound 29 was obtained as a yellow solid (224 mg, 46% for two steps). 1H NMR
(400 MHz, MeOD) δ 8.47 (s, 1H), 8.41 (s, 1H), 8.32 (d, J = 4.8 Hz, 1H), 7.89 (s, 1H),
7.81-7.70 (m, 2H), 7.66 (s, 1H), 7.39 (dd, J = 8.4, 4.9 Hz, 1H), 7.34 (t, J = 7.9 Hz,
1H), 7.24 (d, J = 7.6 Hz, 1H), 7.08 (s, 1H), 4.16 (t, J = 6.2 Hz, 2H), 3.95 (s, 3H), 3.50
(s, 1H), 3.38 (t, J = 7.0 Hz, 2H), 1.94-1.91 (m, 2H), 1.74-1.72 (m, 2H), 1.62-1.60 (m,
13
2H). C NMR (125 MHz, MeOD) δ 157.0, 155.5, 151.8, 149.3, 145.3, 144.5, 138.9,
134.8, 132.2, 128.4, 127.4, 125.7, 124.0, 123.0, 122.7, 108.9, 105.0, 102.0, 82.8, 77.3,
68.9, 55.2, 41.6, 28.6, 28.3, 23.0. HRMS (ESI) calcd m/z for C29H28N8O2 [(M + H)+]
521.2408, found 521.2411. HPLC analysis: 95.6% in purity.
4.2.5.
2-Cyano-1-(7-((4-((3-ethynylphenyl)amino)-7-methoxyquinazolin-6-yl)oxy)heptyl)-3-(
pyridin-4-yl)guanidine (30)
24

7-bromoheptanenitrile (0.14 ml, 0.95 mmol) and potassium carbonate (218 mg,
1.58 mmol) were added to a stirred solution of compound 23 (230 mg, 0.79 mmol) in
DMF (20 ml) and the mixture was stirred for 8 h at rt. DMF was evaporated under
reduced pressure and the residue was extracted with EtOAc (30 ml × 3). The
combined organic layers were washed with brine, dried over anhydrous Na2SO4,
filtered, and concentrated under reduced pressure. The resulting mixture was purified
by silica column chromatograph and eluted with petroleum ether /EtOAc (3 :1) to
give the nitrile that was used in the next step.
The above intermediate nitrile was dissolved in anhydrous THF (100 ml) and then
LiAlH4 (60 mg, 1.58 mmol) was slowly added at 0 °C under an argon atmosphere.
The mixture was stirred for 16 h at rt and then quenched with saturated aqueous
NH4Cl solution and the THF was evaporated under reduced pressure. The residue was
extracted with EtOAc (50 ml × 3). The combined organic layers were washed with
saturated aqueous NaHCO3 solution and brine, dried over anhydrous Na2SO4, filtered,
and concentrated under reduced pressure. The resulting mixture was purified by silica
column chromatography and eluted with petroleum ether/EtOAc (1:1) to give the
amine that was used in the next step.
The above intermediate amine was dissolved in pyridine (20 ml), and then
compound 18 (151 mg, 0.79 mmol), 4-dimethylaminopyridine (9.6 mg, 0.079 mmol),
triethylamine (0.22 ml, 1.58 mmol) were added. After stirring at 50 °C overnight, the
reaction mixture was cooled to rt. The pyridine was evaporated under reduced
pressure and then H2O (20 ml) was added. The aqueous layer was extracted twice
25

with EtOAc (30 ml) and the organic layer washed with brine, dried over anhydrous
Na2SO4, filtered and then concentrated under reduced pressure. The resulting mixture
was purified by silica column chromatography and eluted with CH2Cl2/MeOH (40:1)
1
to give 30 (145 mg, 34% for three steps) as a yellow solid. H NMR (400 MHz,
DMSO-d6) δ 9.47 (s, 1H), 8.49 (s, 1H), 8.37 (d, J = 5.7 Hz, 2H), 7.99 (s, 2H), 7.89 (d,
J = 8.2 Hz, 1H), 7.80 (s, 1H), 7.39 (t, J = 7.9 Hz, 1H), 7.28-7.08 (m, 4H), 4.18 (s, 1H),
4.12 (t, J = 6.5 Hz, 2H), 3.93 (s, 3H), 3.29 (s, 2H), 1.81 (q, J = 7.2, 6.6 Hz, 2H),
1.64-1.29 (m, 8H). 13C NMR (125 MHz, MeOD) δ 156.8, 155.3, 152.2, 149.2, 149.0,
147.0, 145.9, 139.2, 128.4, 127.2, 125.5, 122.8, 122.6, 116.5, 114.9, 109.1, 105.6,
101.8, 82.9, 77.3, 68.9, 55.1, 42.0, 28.8, 28.6, 28.5, 26.2, 25.6. HRMS (ESI) calcd m/z
for C31H32N8O2 [(M + H)+] 549.2721, found 549.2719. HPLC analysis: 95.9% in
purity.
4.2.6.
2-Cyano-1-(7-((4-((3-ethynylphenyl)amino)-7-methoxyquinazolin-6-yl)oxy)heptyl)-3-(
pyridin-3-yl)guanidine (31)
The procedure was the same as that described above for the synthesis of compound
1
30. Compound 31 was obtained as a yellow solid (90 mg, 41% for two steps). H
NMR (400 MHz, MeOD) δ 8.50-8.43 (m, 1H), 8.40 (d, J = 1.8 Hz, 1H), 8.33 (d, J =
4.8 Hz, 1H), 7.89 (d, J = 2.3 Hz, 1H), 7.76 (t, J = 8.9 Hz, 2H), 7.67 (d, J = 1.9 Hz,
1H), 7.41 (dd, J = 8.3, 5.0 Hz, 1H), 7.34 (t, J = 7.9 Hz, 1H), 7.24 (d, J = 7.6 Hz, 1H),
7.11 (s, 1H), 4.15 (t, J = 6.0 Hz, 2H), 3.96 (s, 3H), 3.49 (s, 1H), 3.33 (d, J = 7.3 Hz,
13
2H), 1.91-1.87 (m, 2H), 1.64-1.55 (m, 4H), 1.44-1.41 (m, 4H). C NMR (125 MHz,
26

MeOD) δ 158.6, 157.0, 155.4, 152.3, 149.3, 146.1, 145.4, 144.5, 139.2, 134.8, 132.2,
128.4, 127.2, 125.6, 124.0, 122.9, 122.7, 117.2, 109.2, 105.7, 101.9, 82.9, 77.2, 69.0,
55.1, 41.7, 28.8, 28.6, 28.6, 26.2, 25.7. HRMS (ESI) calcd m/z for C31H32N8O2 [(M +
H)+] 549.2721, found 549.2725. HPLC analysis: 96.3% in purity.
4.2.7.
tert-Butyl-(5-((4-((3-ethynylphenyl)amino)-7-methoxyquinazolin-6-yl)oxy)pentyl)carb
amate (25)
Compound 14 was slowly added at 0 °C to a stirred solution of compound 23 (16.5
g, 56.6 mmol), potassium carbonate (15.6 g, 113.3 mmol) in N, N-dimethylformamide
(100 ml), then the reaction mixture was allowed to warm to rt. After stirring at rt for 8
h, the solvent was evaporated under reduced pressure. The residue was extracted with
EtOAc (80 ml × 3), then the combined organic layers were washed with brine, dried
over anhydrous Na2SO4, filtered and concentrated under reduced pressure. The
residue was purified by silica column chromatography and eluted with petroleum
1
ether/EtOAc (1:1) to give 25 (19.0 g, 68%) as a faint yellow solid. H NMR (400
MHz, CDCl3) δ 8.66 (d, J = 1.8 Hz, 1H), 8.53 (s, 1H), 7.87 (s, 1H), 7.76 (d, J = 8.0
Hz, 1H), 7.47 (s, 1H), 7.35-7.19 (m, 3H), 4.77 (s, 1H), 4.00 (t, J = 7.0 Hz, 2H), 3.94 (s,
3H), 3.14 (q, J = 7.0 Hz, 2H), 3.07 (s, 1H), 1.85-1.81 (m, 2H), 1.53-1.48 (m, 2H),
13
1.46-1.41 (m, 2H), 1.41 (s, 9H). C NMR (125 MHz, CDCl3) δ 162.5, 156.5, 155.0,
153.4, 148.7, 147.2, 139.1, 128.8, 127.6, 125.5, 122.7, 122.6, 109.4, 107.5, 101.6,
83.5, 79.5, 77.2, 69.0, 56.1, 39.9, 29.6, 28.4, 27.7, 22.3.
4.3. Biological assays
27

4.3.1. In vitro NAMPT inhibition assay
The NAMPT enzyme inhibitory activity was evaluated using a CycLex NAMPT
colorimetric assay kit (CycLex NAMPT colorimetric assay kit, MBL International
Corp., Woburn, MA) following the instructions of the manufacturer60. Assay buffers-1
and -2 were prepared before starting the assay. Assay buffer-1 contains: 10 × NAMPT
assay buffer 10 μL, 10 × nicotinamide 10 μL, 10 × PRPP 10 μL, 10 × ATP 10 μL,
recombinant NMNAT1 2 μL, dH2O 48 μL, total volume 90 μL. Assay buffer-2
contains: 50 × WST-1 2 μL, 50 × ADH 2 μL, 50 × diaphorase 2 μL, 10 × EtOH 10 μL,
distilled H2O 4 μL, total volume 20 μL. Recombinant NAMPT (2 μL) and various
concentrations of tested compounds or vehicle were added to each well of the
microplate, and the reaction was initiated by adding 90 μL of assay buffer-1 to each
well and mixing thoroughly followed by incubation at 30 °C for 60 min. Subsequently,
assay buffer-2 (20 μL) was added to each well of the microplate and mixed thoroughly.
The absorbance at 450 nm was monitored for 30 min at 5 min intervals using a
microtiter plate reader.
4.3.2. In vitro EGFR activity assay
According to the instructions of manufacturers, wild type and different EGFR
mutants (T790M, L858R, L861Q, L858R/T790M) activity were tested using the
Z′-Lyte Kinase Assay Kit (Invitrogen).
4.3.3. Cell culture and cell viability assay
Cell lines, Huh-7, MCF7and K562 were purchased from Shanghai Cell Bank,
28

Chinese Academy of Sciences. Cells were routinely grown and maintained in RPMI
or DMEM media with 10% FBS and 1% penicillin/streptomycin. All cell lines were
incubated in a Thermo/Forma Scientific CO2 water jacketed incubator with 5% CO2 in
air at 37 °C. Cell viability assay was determined by the CCK8 (DOjinDo, Japan) assay.
Cells were seeded at a density of 400-800 cells/well in 96-well plates and treated with
various concentration of tested compounds or vehicle. After 72 h incubation, CCK8
reagent was added and absorbance was measured at 450 nm using Envision 2104
multilabel reader (Perkin Elmer, USA). Dose response curves were plotted to
determine the IC50 values using Prism 5.0 (GraphPad Software Inc. USA).
4.3.4. In vivo antitumor efficacy in the H1975 tumor model
All the procedures for animal handling, care, and the treatment in this study were
performed according to the guidelines approved by the Institutional Animal Care and
Use Committee (IACUC) of China Pharmaceutical University following the
guidelines of the Association for Assessment and Accreditation of Laboratory Animal
Care (AAALAC). Female BALB/c nude mice between 4 to 6 weeks of age were
housed in individual HEPA-ventilated cages on a 12 hours light–dark cycle at 21-
23 °C and 40-60% humidity, and used for tumor xenografts. H1975 lung cancer cells
(3 x 106) were subcutaneously implanted in the right flanks of nude mice in a volume
of 0.1 ml of RPMI-1640 medium, and the mice were observed for 4 weeks. 12 days
after the injection of the cells, the xenograft mice were divided into the three group
(nꢀ= 6 per group) with a similar tumor volume, which reached ~150 mm3. Gefitinib
was suspended in 0.9% normal saline and administered once daily (60 mg/kg, i.g.).
Compound 28 was suspended in the solution (including 5% ethanol, 5% dimethyl-
sulfoxide, 10% Cremophor EL and 80% normal saline) and administered five times
every week (60 mg/kg, i.v.). Mice in the vehicle group were given the same solution
as that of compound 28 by intravenous injection. Tumor dimensions were measured
29