High-affinity epidermal growth factor receptor (EGFR) irreversible inhibitors with diminished chemical reactivities as positron emission tomography (PET)-imaging agent candidates of EGFR overexpressing tumors

Article abstract of DOI:10.1021/jm0580196

Previous studies with the anilinoquinazoline epidermal growth factor receptor (EGFR) irreversible inhibitor [¹¹C]-ML03 demonstrated a rapid metabolism of the tracer, which led to its low in vivo accumulation in EGFR overexpressing tumors. To enhance tumor uptake, the chemical structure of the compound was modified, and four new groups of EGFR inhibitors with a wide range of chemical reactivities were synthesized. Chemical reactivity assay of the compounds, performed with reduced glutathione (GSH), revealed that the group C (4-(dimethylamino)-but-2-enoic amide) derivative was the least chemically reactive against the nucleophilic attack of GSH. Nonetheless, it demonstrated a high inhibitory potency and bound irreversibly to the EGFR. Consequently, the blood stability of the group C compound (5a, ML04) labeled with ¹¹C was studied. In a time frame of 60 min, no radioactive metabolites were detected in blood. The stability of [¹¹C]-5a, as indicated both from in vitro blood-stability assays and injection into nude rats, was significantly higher as compared to [¹¹C]-ML03. Since group C presented a greater promise for tumor accumulation, it represents, to date, the most suitable candidate for radiolabeling with long-lived positron emission tomography (PET) radioisotopes.

Full text of DOI:10.1021/jm0580196

J. Med. Chem. 2005, 48, 5337-5348
5337
High-Affinity Epidermal Growth Factor Receptor (EGFR) Irreversible
Inhibitors with Diminished Chemical Reactivities as Positron Emission
Tomography (PET)-Imaging Agent Candidates of EGFR Overexpressing Tumors
Eyal Mishani,*,†,‡ Galith Abourbeh,†,§,‡ Orit Jacobson,† Samar Dissoki,† Revital Ben Daniel,† Yulia Rozen,†
Mazal Shaul,† and Alexander Levitzki§
Department of Medical Biophysics and Nuclear Medicine, Hadassah Hebrew University, Jerusalem 91120, Israel, and
Unit of Cellular Signaling, Department of Biological Chemistry, The Alexander Silberman Institute of Life Sciences,
The Hebrew University, Jerusalem 91904, Israel
Received March 29, 2005
Previous studies with the anilinoquinazoline epidermal growth factor receptor (EGFR)
irreversible inhibitor [11C]-ML03 demonstrated a rapid metabolism of the tracer, which led to
its low in vivo accumulation in EGFR overexpressing tumors. To enhance tumor uptake, the
chemical structure of the compound was modified, and four new groups of EGFR inhibitors
with a wide range of chemical reactivities were synthesized. Chemical reactivity assay of the
compounds, performed with reduced glutathione (GSH), revealed that the group C (4-(di-
methylamino)-but-2-enoic amide) derivative was the least chemically reactive against the
nucleophilic attack of GSH. Nonetheless, it demonstrated a high inhibitory potency and bound
irreversibly to the EGFR. Consequently, the blood stability of the group C compound (5a, ML04)
labeled with 11C was studied. In a time frame of 60 min, no radioactive metabolites were detected
in blood. The stability of [11C]-5a, as indicated both from in vitro blood-stability assays and
injection into nude rats, was significantly higher as compared to [11C]-ML03. Since group C
presented a greater promise for tumor accumulation, it represents, to date, the most suitable
candidate for radiolabeling with long-lived positron emission tomography (PET) radioisotopes.
Introduction
and it contributes to the initiation, progression, and
invasiveness of human cancers.5,7-9 Thus, the EGFR is
an attractive target for the design and development of
compounds that can specifically bind the receptor and
inhibit its tyrosine kinase (TK) activity and its signal
transduction pathway in cancer cells. Such compounds
can serve for therapeutics and, when labeled, as tar-
geted diagnostic agents. The EGFR reversible inhibitor
gefitinib (ZD1839, Iressa; AstraZeneca, London, United
Kingdom) (Figure 1) was approved by the FDA as a
third-line treatment of chemotherapy-refractory non-
small-cell lung carcinomas (NSCLC).10 The anti-EGFR
antibody Erbitux (cetuximab, ImClone systems, Inc.)
has recently gained an accelerated FDA approval as well
for the treatment of metastatic, advanced colorectal
cancer.11 Several other anti-EGFR targeted molecules,
such as erlotinib (OSI-774, Tarceva; Genentech, San
Francisco, CA) (Figure 1), are presently undergoing
phase 3 clinical trials. Consequently, there has been a
considerable interest in the use of radioactively labeled
EGFR inhibitors as bioprobes for positron emission
tomography (PET) molecular imaging of EGFR overex-
pressing tumors.12-17 PET is a nuclear imaging modality
that allows four-dimensional, quantitative determina-
tion of the distribution of radioactive probes within the
human body. It is used for the measurement of physi-
ological, biochemical, and pharmacological functions at
the molecular level, both in normal and in pathological
states. PET allows accurate measurements of radioac-
tivity concentrations in small volume compartments in
vivo, thus enabling kinetic surveillance of radioactivity
throughout the body. PET imaging requires the admin-
Growth factors, differentiation factors, and hormones
are crucial components of the regulatory network that
coordinates the development of multicellular organisms.
Many of these growth factors mediate their pleio-
tropic actions through cell surface receptors with an
intrinsic protein tyrosine kinase activity. The epidermal
growth factor receptor (EGFR/Her-1/ErbB-1), which
belongs to the ErbB receptor family, is involved in the
proliferation and differentiation of both normal and
malignant cells.
The activation of the receptor occurs by a three-step
mechanism: (a) binding of specific ligands such as
epidermal growth factor (EGF), transforming growth
factor R (TGFR), amphiregulin (AR), betacellulin (BTC),
epiregulin, or HB-EGF to the receptor at the extracel-
lular ligand-binding domain; (b) dimerization of the
ligand-receptor complexes; (c) autophosphorylation of
the receptor at the intracellular tyrosine kinase domain.
The autophosphorylation serves as a trigger for several
downstream signal transduction pathways.1-5 Following
the autophosphorylation, the receptor clusters in coated
pits and is endocytosed in vesicles that ultimately either
recycle back to the membrane or fuse with lysosomes,
where the receptor is degraded.6
Overexpression of the EGFR and its enhanced signal-
ing are a frequent hallmark of human epithelial cancers,
* To whom correspondence should be addressed. Fax: 972
6421203. Tel: 972 2 6777931. E-mail: mishani@md.huji.ac.il.
† Hadassah Hebrew University.
2
‡ These authors equally contributed to the article.
§ The Hebrew University.
10.1021/jm0580196 CCC: $30.25 © 2005 American Chemical Society
Published on Web 07/12/2005

5338 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 16
Mishani et al.
Figure 1. Chemical structures of reversible and irreversible EGFR inhibitors.
Figure 2. Chemical structures of group A-D inhibitors of the EGFR.
istration of a suitable molecule labeled with a positron
emitting nuclide, such as oxygen-15, nitrogen-13, carbon-
11, and fluorine-18, which have half-lives of 2.037,
9.965, 20.39, and 109.8 min, respectively.
the labeled compound in the tumor.19 To decrease the
chemical reactivity of the double bond in the acrylamide
moiety and enhance tumor availability, modifications
of the functional group at the 6-position of the quinazo-
line ring were made. We report here on the syntheses
of new EGFR inhibitors (Figure 2), their relative chemi-
cal reactivities, and their inhibitory potencies.
Biological evaluation of labeled anilinoquinazolines
as potential PET-EGFR bioprobes in cancer was previ-
ously reported.18,19 During the course of that study, the
uptake of a labeled reversible inhibitor (MLO1) was
compared with that of a labeled irreversible inhibitor
(MLO3) (Figure 1) in EGFR overexpressing tumors.18-20
Rapid washout of the labeled reversible inhibitor from
the tumor was observed, most likely due to its futile
competition with the high intracellular concentration
of ATP at the ATP-binding pocket of the receptor.18 This
rapid washout was overcome using the irreversible
inhibitor, [11C]-ML03, which covalently binds the EGFR
through the acrylamide substituent on the quinazoline
ring. Nevertheless, the unsaturated, chemically reactive
acrylamide bond of the compound led to its rapid in vivo
metabolism, resulting in only a minor accumulation of
Results
Chemistry. The syntheses of compounds 3-5 and 9
are outlined in Scheme 1. 6-Nitroquinazolone was
reacted with thionyl chloride to produce the 4-chloro-
6-nitroquinazoline. The excess of thionyl chloride was
distilled out, and the product was obtained without any
further purification, with 80% chemical yield and 97%
purity, as determined by HPLC. Reaction of the 4-chloro-
6-nitroquinazoline was performed with either of the
following aniline derivatives: 3,4-dichloro-6-fluoro-
aniline (a), meta-bromoaniline (b) or meta-iodoaniline
(c) in i-PrOH at reflux. The products were obtained as

EGFR Inhibitors with Diminished Chemical Reactivities
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 16 5339
Scheme 1. Chemical Syntheses of Group A-D Inhibitors of EGFR
a bright yellow precipitate. After filtration, compounds
1a-c were obtained with 75%-80% yield. In the next
step, the nitro group at the 6-position of the quinazoline
ring was reduced to amine. It should be noted that an
undesired reduction of the halogen position at the
anilino moiety was observed only with the iodo deriva-
tive. Therefore, the reduction step of this derivative was
performed at 0 °C. Yet, even at this temperature, we
observed 10% formation of the reduction side product
that was separated from the desired product by silica
chromatography. Compounds 2a-c were used as basic
starting materials for the preparation of the three
different groups of inhibitors. They were reacted with
R-methoxyacetyl chloride to yield group A compounds,
3a-c, with 71%, 69%, and 64% yield, respectively.
Reaction with R-chloroacetyl chloride yielded group B
compounds, 4a-c, with 74%, 40%, and 27% yield,
respectively. Group C compounds, 5a-c, were obtained
after a two-step synthesis, as described in Scheme 1,
with 20% yield. The synthesis of group D was started
with reaction of fluoroantrinilic acid and formamide to
obtain the quinazoline (Scheme 1B). Nitration and then
chlorination with thionyl chloride yielded the nitro-
chloroquinazoline. Coupling with the appropriate anili-
no derivative and reaction with methylpiperazine, fol-
lowed by reduction of the nitro group and reaction with
acryloyl chloride provided group D compounds, 9a,b,
with an overall average yield of 11%.
Radiochemistry. Carbon-11-labeled MeI was pre-
pared according to a well-documented procedure and
reacted with 8 mg of 4-monomethyl-amino-but-2-enoic
acid [4-(3,4-dichloro-6-fluoro-phenylamino)-quinazolin-
6-yl]-amide precursor.14 The product, [11C]-5a, was
obtained after a 50 min total radiosynthesis time,
including HPLC purification, with an average (n ) 30)
final dose of approximately 0.74 GBq (20 mCi, 10%
decay-corrected radiochemical yield) and a specific
activity of 2.1-3.3 Ci/µmol.

5340 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 16
Mishani et al.
Figure 3. Reaction profiles of compounds 4a and ML03 with
reduced glutathione. Compounds were dissolved in THF-
MeOH-H2O (1:2:1) and reacted at room temperature with 1/2
equiv of reduced glutathione in the presence of N,N′-diisopro-
pylethylamine. The conversion rate into GSH conjugates was
measured as described in the text.
Figure 5. Calculation of activation energy of ML03: Ea
5.24 kcal mol-1, ∆H2q5°C ) 4.64 kcal mol-1, and ∆Sq25°C
)
)
-61.24 cal mol-1 K-1
.
Figure 4. Reaction profiles of compounds 3a and 5a with
reduced glutathione. Compounds were dissolved in THF-
MeOH-H2O (1:2:1) and reacted at room temperature with 1/2
equiv of reduced glutathione in the presence of N,N′-diisopro-
pylethylamine. The conversion rate into GSH conjugates was
measured as described in the text.
Challenge Reaction with Reduced Glutathione.
EGFR blockade by irreversible inhibitors is due to the
nucleophilic attack of the sulfhydryl group of Cys-773
at the ATP binding pocket of the receptor on the reactive
chemical group of the EGFR-targeted inhibitor.21 To
mimic this reaction, we examined the degree of reactiv-
ity of the inhibitors toward the sulfhydryl group of
reduced glutathione (GSH). Compounds 3-5a and
ML03 were dissolved in THF/MeOH/H2O (1:2:1) and
Figure 6. Calculation of activation energy of compound 4a:
Ea ) 11.4 kcal mol-1, ∆H2q5°C ) 10.80 kcal mol-1, and ∆Sq25°C
-41.29 cal mol-1 K-1
)
.
cal mol-1 K-1; for compound 4a, Ea ) 11.4 kcal mol-1
,
∆Hq25°C ) 10.80 kcal mol-1, and ∆S2q5°C ) -41.29 cal
mol-1 K-1. Activation energies of compounds 3a and 5a
were too high; thus even at temperatures exceeding 100
°C, we could not obtain a change in the reaction rate.
The results of the challenge assay of the different
groups of compounds with reduced glutathione placed
groups A and C as the most chemically stable com-
pounds regarding the nucleophilic attack performed by
the sulfhydryl group of GSH.
1
reacted at room temperature with /2 equiv of reduced
glutathione in the presence of 12 equiv of N,N-diisopro-
pylethylamine. Identical aliquots of the reaction mix-
tures were taken at various time points and injected
into reversed-phase HPLC to determine the conversion
rate of the various compounds into glutathione conju-
gates. The characteristics of the products were deter-
mined using MS. For each of the four reactions, a graph
of product concentration as a function of time was
plotted (Figures 3 and 4). Reaction rate constants of 5
× 10-8, 7.0 × 10-5, 7.0 × 10-7, and 1.0 × 10-4 M/min
were measured for compounds 3-5a and ML03, respec-
tively. In a study of the reaction rate as a function of
the temperature (Figures 5 and 6), activation param-
eters were obtained as follows: for ML03, Ea ) 5.24 kcal
mol-1, ∆Hq25°C ) 4.64 kcal mol-1, and ∆Sq25°C ) -61.24
Biology. IC50 Evaluation Assays. Two different
methods were used to measure EGFR-TK autophospho-
rylation IC50 values of the inhibitors and the reference
inhibitor, ML03. In the first method, the inhibitors were
evaluated in a cell-free system by means of ELISA to
determine their EGFR autophosphorylation IC50 values.
The assays were performed with A431 cell lysates
(human epidermoid vulval carcinoma cells), which over-
express the EGFR. The apparent IC50 averages of each

EGFR Inhibitors with Diminished Chemical Reactivities
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 16 5341
Table 1. IC50 Values in the ELISA Screen and in Intact A431
whereas those of group A and D compounds were about
2-3 orders of magnitude higher.
Cellsa
Intact A431 cells
To assess the irreversible inhibitory effect of the
compounds upon EGFR autophosphorylation, in the
second method, each inhibitor was incubated with intact
A431 cells for 1 h, as described in the Experimental
Section, and the degree of EGFR phosphorylation was
measured either immediately after or 8 h after removal
of the inhibitor from the medium. The results presented
in Figure 7 are summarized in Table 1. In accordance
with the data that were obtained from the cell-free
assays, the studies with intact A431 cells demonstrate
that not only were group B and C compounds more
potent than their analogues of groups A and D in
inhibiting the EGFR but the former also retained
their inhibitory effect on receptor phosphorylation even
8 h after their removal from the medium, affirming
that these inhibitors are covalently bound to the re-
ceptor.
Blood Stability of [11C]-ML03 and [11C]-5a and
in Vivo Metabolism Assessment. It was previously
demonstrated that following incubation of [11C]-ML03
with blood samples, the total extracted radioactivity
decreased from 65% to 50% within 1 h with no apparent
formation of radioactive metabolites.19 To compare the
blood stability of ML03 with that of compound 5a, a
similar assay was performed with the latter. Briefly,
[11C]-5a was incubated with blood samples at 37 °C for
different time periods. Blood samples were extracted
with ethanol/THF (1:1), the extracted radioactivity was
loaded on reversed-phase TLC plates, and radioactive
bands were visualized by a phosphor imager plate. The
average total basal extracted radioactivity by this
method was 61%, all of which corresponded to [11C]-5a.
IC50 rangec
(immediately
after removal
of the
inhibitor)
[nM]
IC50 rangec
(8 h after
removal
of the
inhibitor)
[nM]
A431 lysate
IC50 appb
[nM]
group structure
ML01
ML03
0.208
0.037
252 ( 137
174 ( 135
150 ( 135
3-5
6.7-20
6.7-20
group A
group B
group C
group D
3a
3b
3c
4a
4b
4c
5a
5b
5c
9a
9b
1000-10000 10000-50000
>13000
1000-5000
>100000
∼50000
25-50
∼10
0.07 ( 0.03 5-25
0.4-32
0.11-0.12
<5
4-10
∼30
0.11 ( 0.08 4-10
0.41 ( 0.39 1-10
0.94 ( 0.90 <5
10-50
1-10
<5
6000 ( 5000 ∼250
>500
>500
500-13000 50-250
a The ELISA was performed with A431 cell membranes. IC50
values in intact A431 cells refer to inhibition of EGFR phospho-
rylation immediately after and 8 h after removal of the compound
from the medium. b The IC50 values for each derivative were
obtained from three independent experiments, at least, and are
represented as either mean ( standard deviation or as a range.
c For each compound, at least two different assays with similar
results were performed. Each experiment was performed in
duplicate.
compound were obtained from three independent dose
response curves and are summarized in Table 1. Unlike
the results that were obtained in the GSH assay, the
results of these experiments revealed that group B and
C compounds possessed the highest inhibitory potencies
toward the EGFR and were on a par with ML03. Their
average IC50 values were in the sub-nanomolar range,
Figure 7. Extent of irreversible inhibitory effect of group A-D compounds upon EGFR phosphorylation, as indicated by its
anti-phosphotyrosine (RPY) content. A431 cells were incubated with the investigated compound for 1 h and stimulated with EGF
(20 ng/mL) either immediately after or 8 h after removal of the compound from the medium. Cell lysates (10 µg of total protein)
were loaded onto 8% SDS-PAGE, and EGFR anti-phosphotyrosine content was measured. Blots of one demonstrative compound
(a derivative) from each group (A-D) are presented.

5342 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 16
Mishani et al.
Table 2. IC50 Values of Group A-C Compounds (a Derivative)
for Inhibition of Phosphorylation of HER1, HER2, and PDGFR
Kinase Domains in Intact DHER14, CSH12, and NIH/PDGFR
Cells, Respectively
For further establishment of relative reactivities of
[11C]-ML03 versus [11C]-5a, an in vivo evaluation of
stability was performed. Briefly, around 250 µCi of the
C-11-labeled compounds were injected into rats; at
predetermined time points, the rats were sacrificed,
blood was drawn, livers were excised, and the extracted
radioactivity was loaded onto reversed-phase TLC plates.
The radio-chromatograms of the extracted radioactivity,
which are presented in Figure 8A,B, revealed that while
metabolite formation was observed at a rather early
stage with [11C]-ML03, no metabolite formation was
detected for the C-11-labeled compound 5a 60 min
postinjection. Altogether, the improved chemical stabil-
ity of compound 5a, as compared to ML03, along with
its potent irreversible inhibitory effect of the EGFR, set
it as a more promising candidate for in vivo imaging of
EGFR overexpressing tumors.
Selectivity Assays. Binding selectivity of a PET
probe to its molecular target is a significant determi-
nant of its ability to serve as a high-quality imaging
agent. To characterize the degree of specificity of the
compounds in inhibiting the EGFR, one representa-
tive compound (the a derivative) of groups A-C was
tested out in a cellular assay, similar to the assay
performed with A431 cells. Group D compounds were
excluded from this assay due to their less favorable
profile, namely, high chemical reactivity and poorer
inhibitory potency. In brief, DHER14, CSH12, and
NIH/PDGFR cells expressing EGFR, EGFR-human
epidermal growth factor receptor 2 (HER2) chimera, or
platelet-derived growth factor receptor (PDGFR), re-
spectively, were incubated with the tested inhibitor for
1 h.22-24 Following removal of the inhibitor from the
medium and stimulation with the appropriate growth
factor, the cells were harvested, and the extent of
inhibition was evaluated by measuring the phosphoty-
rosine content of the receptor in a Western blot analysis.
The data presented in Table 2 reveal that all three
groups of compounds bear no inhibitory effect upon the
PDGFR (IC50 > 1 µM). Nonetheless, the inhibitory
profile with respect to the kinase domain of HER2 and
EGFR was similar to that observed in A431cells: groups
B and C were far more potent in inhibiting both EGFR/
c-ErbB1 (IC50 ) 4-15 nM) and c-ErbB2 (IC50 ) 25-50
nM) than group A (IC50 > 0.5 µM). These results
indicate that group B and C inhibitors possess higher
affinities toward the c-ErbB family of receptors than
toward the PDGFR.
IC50 value (nM)
structure
DHER14
CSH12
NIHPDGFR
3a
4a
5a
>250
5-15
∼4
>500
25-50
25-50
>1000
>1000
>1000
Within 60 min of incubation, the total extracted radio-
activity remained identical (data not presented). As with
[11C]-ML03, formation of radioactive metabolites was
not detected within 1 h of measurement by this extrac-
tion method. However, while the total extracted radio-
activity of [11C]-ML03 samples decreased by nearly 25%
within 1 h, it remained unchanged with [11C]-5a. Most
likely the nonextractable fractions could be accounted
for by binding to sulfhydryl groups of plasma proteins.
These results are in good agreement with those obtained
in the GSH assay and reflect the superior chemical
stability of compound 5a over ML03.
Figure 8. In vivo stability of [11C]-ML03 compared to
[11C]-5a. Nude Hsd:RH-rnu/rnu rats were injected with about
250 µCi of either [11C]-ML03 or [11C]-5a and sacrificed at
predetermined time points. The total extracted radioactivity
from the blood and the liver was measured and loaded onto
RP TLC plates next to a standard for detection of radioactive
metabolites. Panel A shows extracted radioactivity from blood
(right panel) and liver (left panel), as compared to the standard
(middle) 30 min postinjection of [11C]-ML03. Panel B shows
extracted radioactivity from blood (left panel) and liver (middle
panel), as compared to the standard (right) 60 min postinjec-
tion of [11C]-5a. Panel C shows a quantitative comparison of
intact tracer extracted from blood following injection of the
labeled tracer into nude rats. For each labeled compound, the
percent extraction of intact tracer was calculated by multiply-
ing the total extracted radioactivity from 1.5 mL blood by the
TLC fraction of the radioactive band, corresponding to the
tracer, as previously described.19
Discussion
In previous studies with the labeled irreversible
inhibitor [11C]-ML03, high metabolic rate and low tumor
availability were observed, resulting in a moderate
tumor to blood uptake ratio of the tracer.19 Thus, to
enhance tumor uptake and improve the quality of
images, our aim was to explore new chemically modified
EGFR irreversible inhibitors, which hold higher stabil-
ity and less reactivity. With this in mind, four new
groups of EGFR inhibitors with varying extents of
chemical reactivities toward nucleophilic attack were
synthesized as candidates for radiolabeling with PET
isotopes. The same basic structure as the acrylami-
doquinazoline derivative, ML03, was used with strategic
changes that were intended to reduce the chemical

EGFR Inhibitors with Diminished Chemical Reactivities
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 16 5343
reactivity of the double bond of the acrylamide moiety
toward nucleophilic attacks. It is highly probable that
interactions of this kind between blood and plasma
components and the inhibitors exist, hampering the
tumor availability of the labeled compound. The chemi-
cal modifications were performed at the 6-position of the
quinazoline ring (Figure 2). The highly chemically
reactive acrylamide group at the 6-position of ML03 was
replaced with either R-methoxyacetamide (group A),
R-chloroacetamide (group B), or 4-dimethylaminocroto-
nylamide (group C) (the last group was first introduced
by Tsou et al.25) In these groups, the R carbon (groups
A and B) or the â carbon (group C) is partially positively
charged and thus sufficiently reactive toward the nu-
cleophilic attack by Cys-773 at the EGFR kinase do-
main. Moreover, it is likely that Cys-773 is ionized to
the S- form, enabling efficient interaction with less reac-
tive electrophiles, such as group A-C inhibitors. The
higher energy gap of the HOMO-LUMO electronic orbi-
tals of the reactive carbon centers of these compounds,
as compared to that of the â carbon in the acryloyl group
of ML03, increases their chemical stability, thereby
augmenting the biological stability of the compounds
and, hopefully, their tumor availability. Thus, if the
inhibitory potencies were not dramatically affected and
remained at the nanomolar range, the new inhibitors
would remain at the receptor binding site long enough
to allow covalent bonding. By retaining the double bond
of the acrylamide intact (group D compounds), we aimed
at preserving the chemical reactivity toward nucleo-
philic attack. Instead, we wished to modify the solubility
of the molecule by adding the methylpiperazine group
at the 7-position of the quinazoline ring. As previously
mentioned, each group contained elements that could
enable radiolabeling with C-11, F-18, iodine, or bromine
PET isotopes. The substantially longer half-life of I-124
and Br-76 may offer prolonged monitoring following
injection of the radiotracer. Such monitoring could be
possible, providing that the labeled tracer stays long
enough in the environment of the receptor. The acryl-
amide moiety at the 6-position of the quinazoline ring
is known to confer to this type of inhibitor a covalent,
irreversible binding nature to the EGFR.21 These data
were supported by performing assays with intact A431
cells in which the long-lasting inhibitory effect of the
compounds was evident even 8 h after removal of the
inhibitors and successive rinse of the media. The stable
interaction between the receptor and the labeled inhibi-
tor guarantees that the tracer is not eliminated from
the tumor cells, thus leaving adequate time for imaging.
described in Scheme 1B, starting with fluoroantranylic
acid, with an overall yield of 11%.
To evaluate the chemical reactivity of the formerly
studied compound, ML03, as well as of the new groups
of compounds (groups A-C), we reacted one derivative
from each group (dichlorofluoroaniline derivative) with
the sulfhydryl-containing tripeptide glutathione as nu-
cleophile and measured the reaction rate constants
(Figures 3 and 4). As was anticipated, ML03 was found
to be the most chemically reactive compound toward the
nucleophilic attack performed by the sulfhydryl group
of GSH, and the addition reaction went on with a
reaction rate constant of 1.0 × 10-4 M/min. The chlo-
roacetamide derivative of group B was found to be less
reactive, possessing a reaction rate constant of 7.0 ×
10-5 M/min. Group A and C derivatives were found to
be far more stable toward the nucleophilic attack of the
sulfhydryl group, possessing reaction rate constants of
5.0 × 10-8 and 7.0 × 10-7 M/min, respectively. More-
over, by determination of the reaction rate as a function
of temperature (Figures 5 and 6), the activation energies
measured for ML03 and compound 4a were 5.24 and
11.4 kcal mol-1, respectively. The considerably higher
activation energies of compounds 3a and 5a, relative
to the group B derivative and ML03, prevented us from
measuring any change in the reaction rate even at
temperatures exceeding 100 °C. On the whole, these
results suggest that group A and C compounds are more
chemically stable as compared to ML03 and group B
compounds.
Since the actual interaction between the inhibitor and
the receptor is comprised of several components and is
not merely based on a nucleophilic attack,21 the new
irreversible inhibitors were evaluated in vitro by means
of ELISA to determine their EGFR autophosphorylation
IC50 values (Table 1). According to this assay, group B
and C compounds, along with ML03, were the most
potent EGFR inhibitors out of the four investigated
groups. The IC50 values of groups B and C, in the sub-
nanomolar range, were 2-3 orders of magnitude lower
than those of groups A and D. On the whole, the more
chemically reactive compounds, 4a and ML03, were also
more potent in inhibiting the receptor, whereas the more
chemically stable compound, 3a, exhibited the lowest
inhibitory potency. Nonetheless, the relatively chemi-
cally stable compound, 5a, according to the GSH assay
was among the most potent inhibitors of the EGFR as
well. To affirm this favorable in vitro profile of minimal
reactivity toward sulfhydryl groups combined with high
EGFR affinity, the extent of inhibition in terms of
reversibility and potency was further investigated in a
cellular assay. It should be noted that the ELISA screen
provided adequate information regarding the relative
inhibitory potencies of the different groups; neverthe-
less, this assay was not sensitive enough to differentiate
between the derivatives in each group. The lack of
highly reproducible differences in IC50 values within
each group prevented us from favoring one particular
substituent. An evaluation of the irreversible inhibitory
effect of the compounds was provided by incubating each
inhibitor with intact A431 cells for 1 h and measuring
the degree of EGFR phosphorylation either immediately
after or 8 h after removal of the inhibitor from the
medium. As previously described,26 an inhibitor could
To prepare the new groups of irreversible EGFR
inhibitors, the chemical route described by Tsou et al.25
and Fry et al.21 was followed with minor modifications
(Scheme 1). The synthesis of groups A-C started with
nitroantranilic acid, and a reaction with formamide,
followed by chlorination, yielded the 4-chloro-6-nitro-
quinazoline. Coupling with the appropriate anilino
derivative, followed by reduction of the nitro group,
provided the basic anilinoquinazoline derivative from
which each group was obtained. Reactions with meth-
oxyacetyl chloride, chloroacetyl chloride, and bromo-
crotonyl chloride (followed by reaction with dimethyl-
amine) yielded groups A, B, and C, respectively. Group
D compounds were obtained by the chemical route

5344 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 16
Mishani et al.
be considered as an irreversible one if 80% inhibition
or more is achieved 8 h after its removal from the
medium, whereas 20-80% inhibition classifies the
compound as “partially irreversible.” According to this
standard, the data in Table 1 indicate that group A
compounds were reversible (80% inhibition after 8 h was
achieved only at concentrations above 10 µM), whereas
group B and C compounds exhibited an irreversible
inhibitory effect upon the EGFR at 200-fold lower
concentrations. Group D compounds, albeit irreversible,
were less potent than the prototype compound, ML03.
These results further support the superior inhibitory
potency of group B and C compounds over group A and
D compounds with regard to the EGFR, despite the
relatively chemically stable nature (as indicated by the
GSH assay) of compound 5a, a demonstrative inhibitor
of group C compounds. The irreversible inhibitory effect
of group B compounds presumably results from covalent
bonding to the kinase domain of the EGFR. This
suggests that a side chain of three atoms attached at
the 6-position of the quinazoline moiety, as in group B
compounds, is sufficient to attain irreversible binding
to the receptor. As a result of these superior chemical
and biological qualities of group C compounds in general
and of derivative 5a in particular, ex vivo and in vivo
stability assessments of the compound were conducted.
For the following radioassays with [11C]-5a, the radio-
labeling approach was based on C-11 methylation
reaction of the monomethylamine group of the amide
attached to the 6-position of the quinazoline ring, using
C-11 methyl iodide as the labeling agent.27 The total
synthesis time was 50 min, and the final average dose
was 0.74 GBq (20 mCi, 10% decay-corrected radio-
chemical yield (n ) 30)); specific activity was 2.1-3.3
Ci/µmol. HPLC analysis of the labeled product solution
revealed high radiochemical (>99%) and chemical puri-
ties.
The stability of [11C]-5a in blood was assessed by
incubation of the labeled compound with blood samples
at 37 °C for different time periods. The extracted
radioactivity was calculated by measurements of total
and nonextractable radioactivities and was loaded on
TLC plates for visualization of radioactive bands by a
phosphor imager plate. The results indicated that the
basal extraction of total radioactivity by this method
was approximately 65%, all of which corresponded to
[11C]-5a. Within 60 min of incubation, no metabolites
were detected on the radio-chromatogram, and the total
extracted radioactivity remained the same. In contrast,
previous studies with [11C]-ML03 indicated that the
percent of extractable [11C]-ML03 from blood samples
throughout 60 min of incubation decreased roughly
by 25%, thus revealing a more durable stability of
[11C]-5a in blood.19
were detected 60 min following injection of [11C]-5a into
rats. In accordance with the in vitro observations, the
blood stability of [11C]-5a in rats exceeded that of [11C]-
ML03 throughout the 1-h experiment. The stability of
[11C]-5a in comparison to [11C]-ML03, as indicated by
the extractable intact tracer from blood 15 min postin-
jection into nude rats, was approximately 50% higher.
The superior stability profile of this tracer is even
further supported given the fact that within 60 min
following injection of the tracers, the total extractable
intact tracer was 3-fold higher in the case of [11C]-5a,
as compared to the previously studied labeled EGFR
irreversible inhibitor [11C]-ML03 (Figure 8C). Finally,
a representative preliminary selectivity examination of
group A-C compounds revealed a good selectivity
profile of group B and C derivatives, indicated by more
than 3-fold higher inhibitory concentrations for the
PDGFR vs the erbB-1 and -2 kinase domains.
Conclusion
To enhance the tumor uptake of irreversible EGFR
inhibitors and to enable improved future in vivo imaging
qualities of labeled EGFR bioprobes, four new groups
of EGFR inhibitors with diminished chemical reactivi-
ties were synthesized as candidates for PET molecular
imaging agents. Enhanced stability characteristics of
group C compounds were demonstrated by chemical
reactivity assays with reduced GSH, as well as by ex
vivo and in vivo blood-stability studies. Altogether with
the highly selective and irreversible inhibition profile
of this group, these data identify group C compounds
as the most suitable candidates for visualization of
EGFR overexpressing tumors.
Experimental Section
General. All chemicals were purchased from Sigma-Aldrich,
Fisher Scientific, Merck, or J. T. Baker. Chemicals were used
as supplied, excluding THF, which was refluxed over sodium
and benzophenone and freshly distilled prior to use. Mass
spectroscopy was performed in EI mode on a Thermo Quest,
Finnigan Trace MS mass spectrometer at the Hadassah-
Hebrew University Mass Spectroscopy Facility. 1H NMR
spectra were obtained on a Bruker AMX 400 MHz instrument.
Elemental analysis was performed at the Hebrew University
Microanalysis Laboratory. Radiosyntheses were carried out on
a [11C]-CH3I module (Nuclear-Interface, Munster, Germany).
Specific radioactivities were determined by HPLC, using cold
mass calibration curves. [11C]-CO2 was produced by the
14N(p,R)11C nuclear reaction on nitrogen containing 1% oxygen,
using an 18/9 IBA-cyclotron. Bombardment was carried out
for 15-30 min with a 26 µA beam of 16 MeV protons. At the
end of bombardment, the target gas was delivered and trapped
by a cryogenic trap in the [11C]-CH3I module.
Primary antibodies were obtained as follows: PY20 anti-
phosphotyrosine (diluted 1:2000) from Santa Cruz Biotech-
nology Inc. 4G10 anti-phosphotyrosine antibody (1:100 dilu-
tion) was produced from Su4G10 hybridoma cells. Horseradish
peroxidase-conjugated anti-mouse IgG (1:10000 dilution) was
obtained from Jackson Immuno Research Growth factors.
Human recombinant EGF and PDGFââ were purchased from
Sigma-Aldrich, Inc.
For in vivo characterization of stability, a similar
assay was performed: rats were injected with the tracer,
blood was drawn, and their livers were excised and
homogenized. As in the blood-stability assay, the total
extracted radioactivity was measured and loaded on
TLC plates. The results of two assays with the previ-
ously studied compound, [11C]-ML03, as well as with
[11C]-5a are presented in Figure 8 for comparison.
Whereas metabolite formation was observed at a rather
early stage with [11C]-ML03, no radioactive metabolites
Cell Culture. NIH3T3 cells transformed with either the
EGFR (DHER14 cells), the HER1-HER2 chimera (CSH12
cells), or the PDGFR (NIH/PDGFR cells) have been de-
scribed.22-24 These and A431 human epidermoid vulval carci-
noma cells were grown in Dulbecco’s modified Eagle’s medium
(DMEM) (Biological industries, Kibbuts Beit Haemek, Israel)
supplemented with 10% fetal calf serum and antibiotics (peni-
cillin 105 units/L, streptomycin 100 mg/L) at 37 °C in 5% CO2.

EGFR Inhibitors with Diminished Chemical Reactivities
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 16 5345
Animal Studies. Nude-Hsd:RH-rnu male rats (11-12
weeks), weighing 250-300 g, were obtained from Harlan
Industries, Inc. All animal studies were conducted under a
protocol approved by the Research Animal Ethics Committee
of the Hebrew University of Jerusalem and in accordance with
its guidelines.
yield pure 3c with 64% yield. Mp ) 159-163°C; MS (m/z)
435.0 (MH)+; 1H NMR (DMSO-d6) δ 10.06 (s, 1H), 9.83 (s,
1H), 8.697 (d, J ) 3 Hz, 1H), 8.571 (s, 1H), 8.277 (dd, J1 ) 5.7
Hz, J2 ) 2.7 Hz, 1H), 7.991 (dd, J1 ) 16.8 Hz, J2 ) 3.3 Hz,
1H), 7.909 (dd, J1 ) 16.5 Hz, J2 ) 0.9 Hz, 1H), 7.781 (d, J )
16.8 Hz, 1H), 7.46 (d, J ) 16.5 Hz, 1H), 7.183 (t, J ) 14.43
Hz, 1H), 4.09 (s, 1H), 3.43 (s, 3H). Anal. (C17H15IN4O2) C, H,
N. HPLC conditions: C18 analytical column, 55% acetate
buffer, pH ) 3.8/45% acetonitrile, flow ) 1 mL/min, Rt )
10.7 min.
N-{4-[(3,4-Dichloro-6-fluoro-phenyl)amino]-quinazoline-
6-yl}-2-chloroacetamide (4a). To a stirred solution of 6-amino-
4-(3,4-dichloro-6-fluoro-phenylamino)quinazoline, 2a (102 mg,
0.315 mmol), in dry THF under nitrogen at 0 °C, N,N-
diisopropylethylamine (134 µL, 0.774 mmol) was added, fol-
lowed by addition of chloroacetyl chloride (62 µL, 0.774 mmol).
The mixture was stirred at 0 °C for 30 min, poured into
saturated NaHCO3, and extracted with EtOAc. The organic
solution was dried with Na2SO4 and evaporated. The crude
material was purified by silica gel chromatography (eluent 3%
MeOH in CH2Cl2) to yield pure 4a with 74% yield. Mp > 300
°C; MS (m/z) 399.0 (MH)+; 1H NMR (DMSO-d6) δ 10.67 (s, 1H),
10.0 (s, 1H), 8.73 (s, 1H), 8.47 (s, 1H), 7.81 (m, 4H), 4.34 (s,
2H). Anal. (C16H10Cl3FN4O) C, H, N. HPLC conditions: C18
analytical column, 55% acetate buffer, pH ) 3.8/45% acetoni-
trile, flow ) 1 mL/min, Rt ) 13.1 min.
N-{4-[(3-Bromo-phenyl)amino]-quinazoline-6-yl}-2-chlo-
roacetamide (4b). This compound was prepared in the same
manner as compound 4a, described above, and was purified
by silica gel chromatography (eluent 3% MeOH in CH2Cl2) to
yield pure 4b with 40% yield. Mp ) 300 °C; MS (m/z) 393
(100%, (MH2)+), 391 (99%, (MH)+); 1H NMR (DMSO-d6) δ 10.63
(s, 1H), 9.98 (s, 1H), 8.71 (s, 1H), 8.58 (s, 1H), 8.14 (bs, 1H),
7.81 (m, 2H), 7.31 (m, 3H), 4.34 (s, 2H). Anal. (C16H12BrClN4O)
C, H, N. HPLC conditions: C18 analytical column, 55% acetate
buffer, pH ) 3.8/45% acetonitrile, flow ) 1 mL/min, Rt ) 13.1
min.
Reactions with Reduced Glutathione. Standard solu-
tions were prepared by dissolving compounds 3-5a and ML03
(0.0146 mmol) in 1.75 mL of THF/MeOH (1:2) and glutathione
(18 mg (0.0586 mmol)) in 0.5 mL of water. A 300 µL aliquot of
the quinazoline standard solution (2.5 µmol) was diluted with
689 µL of THF/MeOH/H2O (1:2:1), after which 11 µL (1.25
µmol) of glutathione solution and 5.22 µL (30 µmol) of N,N′-
diisopropylethylamine were added. Conversions of the quinazo-
lines and formation of conjugates at different time points were
measured by HPLC using RP (3.9 mm × 300 mm) column
(mobile phase, acetonitrile and acetate buffer 0.1 M (2:3) at a
flow rate of 1 mL/min). The glutathione conjugates were
detected by MS as well.
A study of the reaction rate as a function of the temperature
generated the following activation parameters: for ML03, Ea
) 5.24 kcal mol-1, ∆H2q5°C ) 4.64 kcal mol-1, and ∆Sq25°C
)
-61.24 cal mol-1 K-1; for compound 4a, Ea ) 11.4 kcal
mol-1, ∆H2q5°C ) 10.80 kcal mol-1, and ∆Sq25°C ) -41.29 cal
mol-1 K-1
.
For compounds 3a and 5a, the activation energies were too
high; thus even at temperatures exceeding 100°C, no change
in the reaction rate was obtained.
The rate equation is presented as follows:
-d[ML0x]
) kobs[ML0x][GSH]
dt
kobs ) A e-E /(RT)
a
where [ML0x] and [GSH] represent the concentrations of the
tested compound and of glutathione, respectively.
N-{4-[(3-Iodophenyl)amino]-quinazoline-6-yl}-2-chlo-
roacetamide (4c). This compound was prepared in the same
manner as compound 4a, described above, and was purified
by silica gel chromatography (eluent 3% MeOH in CH2Cl2) to
yield pure 4c with 27% yield. Mp > 300 °C; MS (m/z) 439.0
(MH)+; 1H NMR (DMSO-d6) δ 10.6 (s, 1H), 9.9 (s, 1H), 8.71 (s,
1H), 8.57 (s, 1H), 8.25 (m, 1H), 7.82 (m, 2H), 7.45 (d, J ) 7.8
Hz, 1H), 7.19 (m, 2H), 4.34 (s, 2H). Anal. (C16H12IClN4O) C,
H, N. HPLC conditions: C18 analytical column, 55% acetate
buffer, pH ) 3.8/45% acetonitrile, flow ) 1 mL/min, Rt ) 13.1
min.
Chemistry. Compounds 1a-c, 2a-c and 5a-c were al-
ready prepared (ref 26 and references therein). Compound 6b
was also synthesized.28
N-{4-[(3,4-Dichloro-6-fluoro-phenyl)amino]-quinazoline-
6-yl}-2-methoxyacetamide (3a). 6-Amino-4-(4,5-dichloro-2-
fluoro-phenylamino)-quinazoline (2a) (62.4 mg, 0.193 mmol)
and triethylamine (53 µL, 39 mg, 0.386 mmol) were dissolved
in dry THF (20 mL) and cooled to 0 °C. 4-Methoxyacetyl
chloride (35 µL, 42 mg, 0.39 mmol) was added, and the mixture
was stirred at 0 °C for half an hour. The mixture was poured
into saturated NaHCO3 (30 mL) and extracted with ethyl
acetate (3 × 30 mL). The organic solution was dried with
Na2SO4 and evaporated. The crude material was purified by
silica gel chromatography (eluent 4% MeOH in CH2Cl2),
generating pure 3a with 71% yield. Mp ) 204-206 °C; MS
4-[(3,4-Dichloro-6-fluoro-phenyl)amino]-7-fluoro-6-ni-
troquinazoline (6a). 4-Chloro-7-fluoro-6-nitro-quinazoline20
(0.245 g, 1.076 mmol) was dissolved in dichloromethane (7 mL),
and 4,5-dichloro-2-fluoro-phenylamine (0.213 g, 1.18 mmol)
dissolved in 2-propanol (14 mL) was added dropwise. The
reaction mixture was stirred at room temperature for 30 min,
while precipitation occurred. The precipitation was completed
by the addition of hexane (25 mL); the crystals were filtered,
dissolved with methanol, and neutralized with triethylamine.
The orange solution was diluted with water to yield a yellow
precipitate, which was filtered and washed with water. The
yellow crystals were extracted with ethyl acetate and washed
three times with NaHCO3 solution (4%). The organic phase
was dried over sodium sulfate, filtered, and evaporated under
reduced pressure to yield pure 6a with 30% yield. Mp ) 232-
1
(m/z) 395.0 (MH+); H NMR (DMSO-d6) δ 10.13 (s, 1H), 9.94
(s, 1H), 8.73 (s, 1H), 8.48 (s, 1H), 7.98 (d, J ) 3.3 Hz, 1H),
7.93 (d, J ) 3.0 Hz, 1H), 7.9 (d, J ) 11.4 Hz, 1H), 7.79 (d, J )
13.8 Hz, 1H), 4.09 (s, 2H), 3.42 (s, 3H). Anal. (C17H13Cl2FN4O2)
C, H, N. HPLC conditions: C-18 column, 60% acetate buffer,
pH ) 3.8/40% acetonitrile, flow ) 1 mL/min, Rt ) 18.39 min.
N-{4-[(3-Bromo-phenyl)amino]-quinazoline-6-yl}-2-
methoxyacetamide (3b). This compound was prepared in
the same manner as compound 3a, described above, and was
purified by silica gel chromatography (eluent 3% MeOH in
CH2Cl2) to yield pure 3b with 69% yield. Mp ) 190-191 °C;
MS (m/z) 387.0 (MH+); 1H NMR (DMSO-d6) δ10.1 (s, 1H), 9.9
(s, 1H), 8.72 (d, J ) 3.6 Hz, 1H), 8.6 (s, 1H), 8.2 (t, J ) 3.6 Hz,
1H), 8.01 (dd, J1 ) 16 Hz, J2 ) 3.6 Hz, 1H), 7.87 (dt, J1 ) 13
Hz, J2 ) 3.4, 1H), 7.82 (d, J ) 16 Hz, 1H), 7.3 (m, 2H), 4.1 (s,
2H), 3.4 (s, 3H). Anal. (C17H15BrN4O2) C, H, N. HPLC condi-
tions: C-18 column, 60% acetate buffer, pH ) 3.8/40% aceto-
nitrile, flow ) 1 mL/min, Rt ) 13.4 min.
1
235° C; MS (m/z) 371.0 (MH)+; H NMR (DMSO-d6) δ 10.65
(bs, 1H), 9.48 (bs, 1H), 8.63 (s, 1H), 7.84 (m, 3H). HPLC
conditions: C-18 column, 45% acetate buffer pH ) 3.8/55%
acetonitrile, flow ) 1 mL/min, Rt ) 11.03 min.
4-[(3-Bromophenyl)amino]-7-fluoro-6-nitroquinazo-
line (6b). This compound was prepared following a known
procedure.21 Mp ) 198-201 °C; MS (m/z) 363.0 (M + H)+; 1H
NMR (DMSO-d6) δ 10.48 (s, 1H), 9.61 (d, J ) 7.8 Hz, 1H), 8.75
(s, 1H), 8.15 (s, 1H), 7.8 (m, 2H), 7.38 (m, 2H). HPLC
conditions: C-18 column, 60% acetate buffer, pH ) 3.8/40%
acetonitrile, flow ) 1 mL/min, Rt ) 24.48 min.
N-{4-[(3-Iodophenyl)amino]-quinazoline-6-yl}-2-meth-
oxyacetamide (3c). This compound was prepared in the same
manner as compound 3a, described above, and was purified
by silica gel chromatography (eluent 3% MeOH in CH2Cl2) to

5346 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 16
Mishani et al.
4-[(3,4-Dichloro-6-fluoro-phenyl)amino]-7-methylpip-
erazine-6-nitroquinazoline (7a). Compound 6a (0.193 g,
0.52 mmol) was dissolved in dry THF (48 mL), and methylpip-
erazine (2.08 mmol, 230 µL) was added. The reaction mixture
was refluxed for 20 h. THF was evaporated under reduced
pressure; the residue was extracted with ethyl acetate and
washed three times with a NaHCO3 solution (4%). The organic
phase was dried over Na2SO4, filtered, and evaporated, yield-
ing pure 7a with 91% yield. Mp ) 196-199 °C; MS (m/z) 453.0
(MH)+; 1H NMR (DMSO-d6) δ 8.95 (s, 1H), 8.4 (s, 1H), 7.85 (d,
J ) 7.5 Hz, 1H), 7.75 (d, J ) 9.9 Hz, 1H), 7.22 (s, 1H), 3.06
(m, 4H), 2.43 (m, 4H), 2.22 (s, 3H). HPLC conditions: C-18
column, 45% acetate buffer, pH ) 3.8/55% acetonitrile, flow
)1 mL/min, Rt ) 11 min.
4-(3-Bromophenylamino)-7-methlypiperazine-6-nitro-
quinazoline (7b). This compound was prepared in the same
manner as compound 7a, described above, to yield pure 7b
with 94% yield. Mp ) 219-222 °C; MS (m/z) 445.0 (MH)+; 1H
NMR (DMSO-d6) δ 8.75 (s, 1H), 8.37 (s, 1H), 8.08 (m, 1H), 7.6
(m, 1H), 7.4 (s, 1H), 7.3 (m, 2H), 3.2 (t, J ) 4.8 Hz, 4H), 2.6 (t,
J ) 4.8 Hz, 4H), 2.38 (s, 3H). HPLC conditions: C-18 column,
60% acetate buffer, pH ) 3.8/40% acetonitrile, flow ) 1
mL/min, Rt ) 16.53 min.
4-(3,4-Dichloro-6-fluoro-phenylamino)-7-methylpiper-
azine-6-aminoquinazoline (8a). Compound 7a (0.212 g,
0.469 mmol) was dissolved in 9:1 ethanol/water (62 mL) at
room temperature, and after the solution was heated to 60
°C, hydrazine monohydrate (0.187 mmol, 91µL) and Ra-Ni
(400 µL) were added. The reaction mixture was heated to 80
°C for 25 min. The brown-orange solution was cooled and
filtered over cellite; the filtrate was evaporated under reduced
pressure and dried to give pure 8a with 91% yield. Mp )
221-225 °C; MS (m/z) 421.1 (MH)+; 1H NMR (DMSO-d6) δ 9.28
(s, 1H), 8.26 (s, 1H), 7.97 (d, J ) 7.8 Hz, 1H), 7.75 (d, J ) 10.2
Hz, 1H), 7.32 (s, 1H), 7.17 (s, 1H), 5.21 (bs, 2H), 2.97 (bs, 4H),
2.54 (bs, 4H), 2.24 (s, 3H). HPLC conditions: C-18 column,
60% acetate buffer, pH ) 3.8/40% acetonitrile, flow ) 1
mL/min, Rt ) 11.48 min.
4-(3-Bromophenylamino)-7-methlypiperazine-6-ami-
noquinazoline (8b). This compound was prepared in the
same manner as compound 8a, described above, and was
purified by silica gel chromatography (eluent 5% MeOH in
CH2Cl2) to yield pure 8b with 55% yield. Mp ) 93-95 °C; MS
(m/z) 415.1 (MH)+; 1H NMR (DMSO-d6) δ 9.39 (s, 1H), 8.37 (s,
1H), 8.22 (m, 1H, m), 7.84 (m, 1H), 7.47 (1H), 7.23 (2H), 7.15
(m, 1H), 5.13 (m, 2H), 2.97 (bs, 4H), 2.55 (bs, 4H), 2.24 (s, 3H).
HPLC conditions: C-18 column, 60% acetate buffer, pH )
3.8/40% acetonitrile, flow ) 1 mL/min, Rt ) 7.5 min.
(s, 1H), 8.69 (s, 1H), 8.54 (bs, 1H), 8.16 (bs, 1H), 7.87 (d, J )
8.1 Hz, 1H), 7.29 (m, 3H), 6.65 (m, 1H), 6.31 (d, J ) 16.8 Hz,
1H), 5.8 (d, J ) 9.9 Hz, 1H), 2.98 (bs, 4H), 2.54 (bs, 4H), 2.24
(s, 3H). Anal. (C22H23BrN6O‚2MeOH) C, H, N. HPLC condi-
tions: C-18 column, 60% acetate buffer, pH ) 3.8/40% aceto-
nitrile, flow ) 1 mL/min, Rt ) 11.98 min.
Radiochemistry. Radiolabeling of Compound 5a. The
radiosynthesis of the carbon-11-labeled compound 5a was
previously described.14 Briefly, [11C]-MeI was prepared accord-
ing to a well-documented procedure and reacted with 8 mg of
4-monomethyl-amino-but-2-enoic acid [4-(3,4-dichloro-6-fluoro-
phenylamino)-quinazolin-6-yl]-amide precursor in a mixture
of 0.3 mL of anhydrous THF, 0.1 mL of CH3CN, and 0.2 mL of
DMSO at -20 °C. At the end of the 2-min distillation step, an
average of 550 mCi (n ) 30) was trapped in the reactor. The
reactor was sealed and heated to 100 °C for 5 min. At the end
of the 5-min reaction, THF and CH3CN were removed under
a flow of argon at 80 °C. The mixture was cooled to 40 °C, 0.6
mL of HPLC solvent (2% THF/60% CH3CN/38% 0.1 M am-
monium formate) was added, and the crude product [average
of 370 mCi (n ) 30)] was automatically injected to the HPLC
[Bischoff Nucleosil 100-7-C18 reverse phase preparative col-
umn (7 µm, 250 mm × 16 mm), flow rate of 13 mL/min]. The
labeled product was collected (Rt ) 18 min) in a flask
containing 60 µL of 1 M NaOH in 85 mL of water. The solution
was passed through a C-18 cartridge (Waters Sep-Pak Plus,
preactivated with 10 mL EtOH and 20 mL of sterile water),
and the cartridge was washed with 4 mL of sterile water. The
product was eluted with 0.75 mL of EtOH, followed by 4.25
mL of saline, and collected into the product vial after a total
radiosynthesis time of 50 min. Identification of the product
and chemical and radiochemical purities were determined by
reversed phase HPLC C18 analytical column. The average (n
) 30) final dose was approximately 0.74 GBq (20 mCi, 10%
decay-corrected radiochemical yield); specific activity was
2.1-3.3 Ci/µmol.
C-11 radiolabeling of ML03 was described earlier.20
Biology. Production of A431 Cell Lysates as an EGFR
Source. A431 human epidermoid vulval carcinoma cells,
which overexpress the EGFR, were used for the production of
active EGFR-containing membranes, as previously described.26
Cell-Free Assay. An evaluation of the EGFR autophos-
phorylation IC50 values was obtained by means of an ELISA
screen, performed as depicted earlier.26 For each compound,
at least three independent assays that gave similar results
were considered. Each assay was carried out using duplicate
samples.
Cellular Assays. Irreversibility Test Protocol. The
inhibitory potency of the different compounds, as well as the
extent of irreversible EGFR inhibition, was evaluated in intact
A431 cells, as already published.26 For each compound, at least
two different assays with similar results were considered. Each
experiment was performed in duplicate.
Selective-Inhibition Assay. CSH12, DHER14, and NIH-
PDGFR cells, expressing either the HER1-HER2 chimera,
EGFR, or PDGFR, respectively, were used for the determina-
tion of inhibitory selectivity.22-24 Cells (7.5 × 104) were grown
in 6-well plates (35 mm diameter, Nalge Nunc) for 24 h and
then incubated in 0.25% FCS-containing medium for an
additional 24 h to ∼90% confluence. Duplicate sets of cells were
treated with the tested compounds at varying concentrations
for 1 h. The final concentration of the vehicle in the medium
was 0.05% DMSO, 0.1% EtOH. After removal of the inhibitor
from the medium, PBS wash (×2), and addition of 0.25%
serum-containing medium to the wells, the cells were stimu-
lated with either 20 ng/mL human EGF for 5 min (CSH12 and
DHER14 cells) or 50 ng/mL human PDGFââ for 10 min
(NIHPDGFR cells) at 37°C. Following the stimulation with the
growth factor, the cells were washed with cold PBS. Cell
extracts were made by adding 0.4 mL of boiling Laemmli buffer
(10% glycerol, 2% sodium dodecyl sulfate, 5% â-mercapto-
ethanol, 62.5 mM Tris‚HCl, pH 6.8) containing 0.001% bro-
mophenol blue to the cells, scraping them with a rubber
policeman and heating to 100 °C for 10 min. For each
6-Acrylamido-7-methylpiperazine-4-[(3,4-dichloro-6-
fluoro-phenyl)amino]-quinazoline (9a). Compound 8a (0.08
g, 0.19 mmol) was dissolved in dry THF (15 mL) and stirred
in an ice bath for 20 min. Diisopropylethylamine (0.132 mL,
0.76 mmol) was added; after 5 min, acryloyl chloride (0.031
mL, 0.38 mmol) was added, and the mixture was stirred for
another 5 min at 0-5 °C. The reaction mixture was left to stir
at room temperature for an additional 1 h under inert
conditions. THF was evaporated; the residue was extracted
with ethyl acetate and washed three times with a NaHCO3
solution (4%). The organic phase was dried over Na2SO4,
filtered, and evaporated. The crude material was purified by
silica gel chromatography (eluent 5% MeOH in CH2Cl2) to give
pure 9a with 44% yield. Mp ) 160-163 °C; MS (m/z) 475.0
1
(MH)+; H NMR (DMSO-d6) δ 9.47 (s, 1H), 8.66 (s, 1H), 8.37
(bs, 1H), 7.79 (m, 2H), 7.28 (s, 1H), 6.68 (m, 1H), 6.3 (d, J )
16.8 Hz, 1H), 5.8 (d, J ) 9.9 Hz, 1H), 2.98 (bs, 4H), 2.55 (bs,
4H), 2.24 (s, 3H). Anal. (C22H21Cl2FN6O‚MeOH) C, H, N. HPLC
conditions: C-18 column, 60% acetate buffer, pH ) 3.8/40%
acetonitrile, flow ) 1 mL/min, Rt ) 15.75 min.
6-Acrylamido-7-methylpiperazine-4-[(3-bromophenyl)-
amino]- quinazoline (9b). This compound was prepared in
the same manner as compound 9a, described above, and
purified by silica gel chromatography (eluent 5% MeOH in
CH2Cl2) to yield pure 9b with 62% yield. Mp ) 125-127 °C.
MS (m/z) 469 (MH)+; 1H NMR (DMSO-d6) δ 9.79 (s, 1H), 9.54

EGFR Inhibitors with Diminished Chemical Reactivities
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 16 5347
(5) Voldborg, B. R.; Damstrup, L.; Spang-Thomsen, M.; Poulsen, H.
S. Epidermal growth factor receptor (EGFR) and EGFR muta-
tions, function and possible role in clinical trials. Ann. Oncol.
1997, 8, 1197-1206.
(6) Stoscheck, C. M.; Carpenter, G. Characterization of the metabolic
turnover of epidermal growth factor receptor protein in A-431
cells. J. Cell. Physiol. 1984, 120, 296-302.
(7) Ritter, C. A.; Artega, C. L. The epidermal growth factor receptor-
tyrosine kinase: a promising therapeutic target in solid tumors.
Semin. Oncol. 2003, 30, 993-1011.
(8) Raymond, E.; Faivre, S.; Armand, J. P. Epidermal growth factor
receptor tyrosine kinase as a target for anticancer therapy.
Drugs 2000, 60, 15-23; discussion 41-42.
compound, at least two different assays with similar results
were performed. Each experiment was carried out in duplicate.
Western Blot Analysis. Ten micrograms of protein lysates
were separated by SDS-PAGE (8% polyacrylamide) and
electrophoretically transferred to a nitrocellulose membrane.
The membrane was blocked in TBST buffer (10 mM Tris-HCl,
pH 7.4 (TBS), 0.2% Tween 20, 170 mM NaCl) containing 5%
low-fat milk for 30 min, then probed overnight with the
appropriate primary antibody. The membrane was washed
thoroughly with TBST and incubated for 1 h with a horserad-
ish peroxidase-conjugated anti-mouse IgG. Finally, the mem-
brane was washed in TBST (4 × 5 min), and immunoreactive
proteins were visualized using an enhanced chemilumines-
cence (ECL) detection reagent.
(9) Salomon, D. S.; Brandt, R.; Ciardiello, F.; Normanno, N.
Epidermal growth factor-related peptides and their receptors in
human malignancies. Crit. Rev. Oncol. Hematol. 1995, 19, 183-
232.
Blood Stability Assay. Blood samples were divided into 1
mL aliquots inside glass vials and incubated for 15 min at 37
°C under gentle shaking. A total of 100-150 µCi (25-50 µL)
of [11C]-5a solution (SA 150-220 µCi/nmol) was added to the
blood samples for predetermined time periods, after which 1.5
mL of EtOH/THF (1:1) was added, and the blood samples were
vortexed and centrifuged (5 min, 3000 rpm) for extraction of
radioactive material. The entire vials were placed in a γ-counter
(1480 Wizard 3”) for measurement of total radioactivity; then
the upper organic phase of the samples was loaded on TLC
plates (Partisil LKC18F silica gel, 5 × 20 cm, 250 µm layer
thickness, Whatman), and the vials containing the nonextract-
able phase were placed once more in the γ-counter for
measurement of the nonextractable radioactivity. The total
extracted radioactivity was calculated by subtracting the
nonextractable radioactivity from the total one. TLC plates
were run for 50 min with 5% THF, 5% ammonium formate
(0.1 M), 90% EtOH as a mobile phase and exposed for 1 h to
phosphor imager plates (BAS-IP MS 2040 Fuji Photo Film Co.,
LTD) for visualization of radioactive bands. The plates were
scanned with a BAS reader 3.1 version scanner and analyzed
with TINA 2.10 g software.
Metabolite Study. Nude Hsd:RH-rnu/rnu male rats (200-
300 g) were injected i.v. with [11C]-5a (specific activity range
of 200-500 µCi/µmol at time of injection, average volume of
211 ( 59 µL, and activity range of 130-550 µCi) and sacrificed
at predetermined time points. Blood (1.5 mL) was drawn from
the heart using a heparinized syringe and immediately mixed
with 7.5 mL of 10% THF/ether for extraction of radioactivity.
The liver (∼2 g) was excised, minced, and homogenized with
4 mL of saline in a tissue grinder (Fenbroek). The homogenate
was mixed with 10 mL of 10% THF/ether for extraction of
radioactive material. The total radioactivity of the samples was
measured in a γ-counter and extracted as previously de-
scribed,19 and the nonextractable radioactivity was measured
again in the γ-counter for further calculations. The organic
extracted phase was loaded onto TLC plates with 5% THF,
10% NaCl (0.1 M) solution, 85% MeOH as mobile phase. As
for the blood-stability assay, radioactive bands were visualized
by a phosphor imager plate.
(10) Cohen, M. H.; Williams G. A.; Sridhara, R.; Chen, G.; Pazdur,
R. FDA drug approval summary: Gefitinib (ZD 1839) (Iressa)
tablets. The Oncologist 2003, 8, 303-306.
(11) FDA
News.
http://www.fda.gov/bbs/topics/NEWS/2004/
NEW01024.html.
(12) Mishani, E.; Bonasera, T. A.; Rozen, Y.; Ortu, G.; Gazit, A.;
Levitzki, A. Fluorinated EGFR-TK inhibitors-based tracers for
PET. J. Labelled Compd. Radiopharm. 1999, 42, S27-S29.
(13) Mishani, E.; Abourbeh, G.; Ortu, G.; Rozen, Y.; Ben-David, I.;
Froimovaci, S.; Dissoki, S.; Gazit, A.; Levitzki, A. Novel EGFR
irreversible tyrosine kinase inhibitor candidates for the diag-
nostic and therapeutic treatment of cancer, J. Labelled Compd.
Radiopharm. 2003, 46, S115.
(14) Mishani, E.; Abourbeh, G.; Rozen, Y.; Laki, D.; Levitzki, A.
Carbon-11 labeling of 4-(dimethylamino)-but-2-enoic acid [4-
(phenylamino)-quinazoline-6-yl]-amides. A new class of EGFR
irreversible inhibitors. J. Labelled Compd. Radiopharm. 2003,
46, S2.
(15) VanBrocklin, H. F.; Dorff, P. N.; Vasdev, N.; O’Neil, J. P.;
Nandanan, E.; Gibbs, A. R. Synthesis of 2′-, 3′- and 4′-[F18]-
fluoroanilinoquinazoline. J. Labelled Compd. Radiopharm. 2003,
46, S117.
(16) Ackermann, U.; Tochon-Danguy, H.; Nice, E.; Nerrie, M.; Young,
K.; Sachinidis, J.; Scott, A. Synthesis, radiolabeling and biologi-
cal evaluation of derivatives of the P210BCR-ABL tyrosine
kinase inhibitor AG957, J. Labelled Compd. Radiopharm. 2003,
46, S118.
(17) Kesarios, N.; Tochon-Danguy, H.; Ackermann, U.; Sachinidis,
J.; Lambert, J.; Nerrie, M.; Nice, E.; Scott, A. Synthesis and
radiolabeling of a tyrosine kinase inhibitor of erbB2/neu (HER2)
receptors. J. Labelled Compd. Radiopharm. 2003, 46, S119.
(18) Bonasera, T. A.; Ortu, G.; Rozen, Y.; Krais, R.; Freedman, N.
M.; Chisin, R.; Gazit, A.; Levitzki, A.; Mishani, E. Potential
(18)F-labeled biomarkers for epidermal growth factor receptor
tyrosine kinase. Nucl. Med. Biol. 2001, 28, 359-374.
(19) Ortu, G.; Ben-David, I.; Rozen, Y.; Freedman, N. M. T.; Chisin,
R.; Levitzki, A.; Mishani, E. Labeled EGFRTK irreversible
inhibitor (ML03). In vitro and in vivo properties, potential as
PET biomarker for cancer and feasibility as anticancer drug.
Int. J. Cancer 2002, 101 (4), 360-370.
(20) Ben-David, I.; Rozen, Y.; Ortu, G.; Mishani, E. Radiosynthesis
of ML03, a novel positron emission tomography biomarker for
targeting epidermal growth factor receptor via the labeling
synthon: [C-11]Acryloyl chloride, Appl. Radiat. Isot. 2003, 58,
209-217.
(21) Fry, D. W.; Bridges, A. J.; Denny, W. A.; Doherty, A.; Greis, K.
D.; Hicks, J. L.; Hook, K. E.; Keller, P. R.; Leopold, W. R.; Loo,
J. A.; McNamara, D. J.; Nelson, J. M.; Sherwood, V.; Smaill, J.
B.; Trumpp-Kallmeyer, S.; Dobrusin, E. M. Specific, irreversible
inactivation of the epidermal growth factor receptor and erbB2,
by a new class of tyrosine kinase inhibitor. Proc. Natl. Acad.
Sci. U.S.A. 1998, 95, 12022-12027.
Acknowledgment. Support was provided by TK-
Signal, Ltd., Jerusalem, Israel, and by Rotem Industries
Ltd., Be’er Sheva, Israel.
(22) Lee, J.; Dull, T. J.; Lax, I.; Schlessinger, J.; Ullrich, A. HER2
cytoplasmic domain generates normal mitogenic and transform-
ing signals in a chimeric receptor. EMBO J. 1989, 8, 167-73.
(23) Honegger, A.; Dull, T. J.; Szapary, D.; Komoriya, A.; Kris, R.;
Ullrich, A.; Schlessinger, J. Kinetic parameters of the protein
tyrosine kinase activity of EGF-receptor mutants with individu-
ally altered autophosphorylation sites. EMBO J. 1988, 7, 3053-
3060.
(24) Shawver, L. K.; Schwartz, D. P.; Mann, E.; Chen, H.; Tsai, J.;
Chu, L.; Taylorson, L.; Longhi, M.; Meredith, S.; Germain, L.;
Jacobs, J. S.; Tang, C.; Ullrich, A.; Berens, M. E.; Hersh, E.;
McMahon, G.; Hirth, K. P.; Powell, T. J. Inhibition of platelet-
derived growth factor-mediated signal transduction and tumor
growth by N-[4-(trifluoromethyl)-phenyl]5-methylisoxazole-4-
carboxamide. Clin. Cancer Res. 1997, 3 (7), 1167-77.
(25) Tsou, H. R.; Mamuya, N.; Johnson, B. D.; Reich, M. F.; Gruber,
B. C.; Ye, F.; Nilakantan, R.; Shen, R.; Discafani, C.; DeBlanc,
R.; Davis, R.; Koehn, F. E.; Greenberger, L. M.; Wang, Y. F.;
Wissner, A. 6-Substituted-4-(3-bromophenylamino)quinazolines
Supporting Information Available: Elemental analysis
of group A-D inhibitors of EGFR. This material is available
free of charge via the Internet at http://pubs.acs.org.
References
(1) Levitzki, A.; Gazit, A. Tyrosine kinase inhibition: an approach
to drug development. Science 1995, 267, 1782-1788.
(2) Yarden, Y.; Schlessinger, J. Self-phosphorylation of epidermal
growth factor receptor: evidence for a model of intermolecular
allosteric activation. Biochemistry 1987, 26, 1434-1442.
(3) Cohen, S.; Ushiro, H.; Stoscheck, C.; Chinkers, M. A native 170,-
000 epidermal growth factor receptor-kinase complex from shed
plasma membrane vesicles. J. Biol. Chem. 1982, 257, 1523-
1531.
(4) Iwashita, S.; Kobayashi, M. Signal transduction system for
growth factor receptors associated with tyrosine kinase activ-
ity: epidermal growth factor receptor signaling and its regula-
tion. Cell. Signalling 1992, 4, 123-132.

5348 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 16
Mishani et al.
as putative irreversible inhibitors of the epidermal growth factor
receptor (EGFR) and human epidermal growth factor receptor
(HER-2) tyrosine kinases with enhanced antitumor activity. J.
Med. Chem. 2001, 44, 2719-2734.
(27) Mishani, E.; Abourbeh, G.; Rozen, Y.; Jacobson, O.; Laky, D.;
Ben David, I.; Levitzki, A.; Shaul, M. Novel carbon-11 labeled
4-(dimethylamino)-but-2-enoic acid [4-(phenylamino)-quinazo-
line-6-yl]-amides: potential PET bioprobes for molecular imaging
of EGFR-positive tumors. Nucl. Med. Biol. 2004, 31, 469-476.
(28) Rewcastle, G. W.; Palmer, B. D.; Bridges, A. J.; Showalter, H.
D.; Sun, L.; Nelson, J.; McMichael, A.; Kraker, A. J.; Fry, D.
W.; Denny, W. A. Tyrosine kinase inhibitors. 9. Synthesis and
evaluation of fused tricyclic quinazoline analogues as ATP site
inhibitors of the tyrosine kinase activity of the epidermal growth
factor receptor. J. Med. Chem. 1996, 39, 918-928.
(26) Smaill, J. B.; Palmer, B. D.; Rewcastle, G. W.; Denny, W. A.;
McNamara, D. J.; Dobrusin, E. M.; Bridges, A. J.; Zhou, H.;
Showalter, H. D.; Winters, R. T.; Leopold, W. R.; Fry, D. W.;
Nelson, J. M.; Slintak, V.; Elliot, W. L.; Roberts, B. J.; Vincent,
P. W.; Patmore, S. J. Tyrosine kinase inhibitors. 15. 4-(Phenyl-
amino)quinazoline and 4-(phenylamino)pyrido[d]pyrimidine acryl-
amides as irreversible inhibitors of the ATP binding site of the
epidermal growth factor receptor. J. Med. Chem. 1999, 42, 1803-
1815.
JM0580196