Design, synthesis and biological evaluation of novel substituted purine isosters as EGFR kinase inhibitors, with promising pharmacokinetic profile and in vivo efficacy

Article abstract of DOI:10.1016/j.ejmech.2019.05.029

Novel substituted purine isosters, were designed and synthesized as potential inhibitors of the Epidermal Growth Factor Receptor (EGFR). The compounds were rationally designed through bioisosteric replacement of the central quinazoline core of lapatinib, an approved drug that inhibits both EGFR and HER2, another important member of this family of receptors. The new target molecules were evaluated as inhibitors of receptor phosphorylation at the cellular level, for their direct inhibitory action on the intracellular receptor kinase domain and for their cytotoxicity against the non-small cell lung cancer cell line A549 and breast cancer HCC1954, cell lines which are associated with overexpression of EGFR and HER2, respectively. The most potent derivatives were further studied for their cellular uptake levels and in vivo pharmacokinetic properties. One compound (23)displayed a noteworthy pharmacokinetic profile, and higher intracellular accumulation in comparison to lapatinib in the A549 cells, possibly due to its higher lipophilicity. This lead compound (23)was assessed for its efficacy in an EGFR positive xenograft model, where it successfully inhibited tumor growth, with a similar efficacy with that of lapatinib and with minimal phenotypic toxicity.

Full text of DOI:10.1016/j.ejmech.2019.05.029

Accepted Manuscript
Design, synthesis and biological evaluation of novel substituted purine isosters as
EGFR kinase inhibitors, with promising pharmacokinetic profile and in vivo efficacy
Efthymios-Spyridon Gavriil, Aris Doukatas, Theodoros Karampelas, Vassilios
Myrianthopoulos, Spyridon Dimitrakis, Emmanuel Mikros, Panagiotis Marakos,
Constantin Tamvakopoulos, Nicole Pouli
PII:
S0223-5234(19)30439-8
DOI:
https://doi.org/10.1016/j.ejmech.2019.05.029
EJMECH 11339
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 29 March 2019
Revised Date: 7 May 2019
Accepted Date: 7 May 2019
Please cite this article as: E.-S. Gavriil, A. Doukatas, T. Karampelas, V. Myrianthopoulos, S. Dimitrakis,
E. Mikros, P. Marakos, C. Tamvakopoulos, N. Pouli, Design, synthesis and biological evaluation of
novel substituted purine isosters as EGFR kinase inhibitors, with promising pharmacokinetic profile
and in vivo efficacy, European Journal of Medicinal Chemistry (2019), doi: https://doi.org/10.1016/
j.ejmech.2019.05.029.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT
Design, Synthesis and Biological Evaluation of Novel Substituted Purine Isosters
as EGFR kinase inhibitors, with promising pharmacokinetic profile and in vivo
efficacy.
Efthymios-Spyridon Gavriil,1§ Aris Doukatas,2§ Theodoros Karampelas,2 Vassilios
Myrianthopoulos,1 Spyridon Dimitrakis,1 Emmanuel Mikros,1 Panagiotis Marakos,1
Constantin Tamvakopoulos,2* Nicole Pouli1*
1 Division of Pharmaceutical Chemistry, Department of Pharmacy, School of Health
Sciences, National and Kapodistrian University of Athens, Panepistimiopolis
Zografou 15771, Athens, Greece
2
Center of Clinical Research, Experimental Surgery and Translational Research,
Biomedical Research Foundation Academy of Athens, Athens, Greece
Author Contributions
§The authors contributed equally.
*Correspondence to:
Nicole Pouli, e-mail: pouli@pharm.uoa.gr
Constantin Tamvakopoulos, e-mail: ctamvakop@bioacademy.gr
ABSTRACT
Novel substituted purine isosters, were designed and synthesized as potential
inhibitors of the Epidermal Growth Factor Receptor (EGFR). The compounds were
rationally designed through bioisosteric replacement of the central quinazoline core of
lapatinib, an approved drug that inhibits both EGFR and HER2, another important
member of this family of receptors. The new target molecules were evaluated as
inhibitors of receptor phosphorylation at the cellular level, for their direct inhibitory
action on the intracellular receptor kinase domain and for their cytotoxicity against the
1

ACCEPTED MANUSCRIPT
non-small cell lung cancer cell line A549 and breast cancer HCC1954, cell lines
which are associated with overexpression of EGFR and HER2, respectively. The most
potent derivatives were further studied for their cellular uptake levels and in vivo
pharmacokinetic properties. One compound (23) displayed a noteworthy
pharmacokinetic profile, and higher intracellular accumulation in comparison to
lapatinib in the A549 cells, possibly due to its higher lipophilicity. This lead
compound (23) was assessed for its efficacy in an EGFR positive xenograft model,
where it successfully inhibited tumor growth, with a similar efficacy with that of
lapatinib and with minimal phenotypic toxicity.
Keywords: lapatinib, purine analogues, EGFR inhibitors, non-small cell lung cancer,
breast cancer, pharmacokinetic studies
1. Introduction
The human epidermal growth factor family of cell surface receptors consists of four
members (HER1-4, ErbB1-4), which serve as main mediators of cell signaling by
extracellular growth factors and play key regulatory roles in essential cellular
functions [1, 2]. They share the typical receptor tyrosine kinase architecture, with an
extracellular ligand binding region, a single transmembrane segment and an
intracellular tyrosine kinase domain [3]. Upon ligand binding, the receptors form
homo and heterodimers and undergo a conformational change that activates the kinase
domain, resulting in trans autophosphorylation and initiation of signal transduction
cascades [4-7]. The abnormal overexpression or the presence of activating mutations
of the HER family members cause malignant cell proliferation [8] and are implicated
in a number of the most lethal solid tumors, such as non-small cell lung cancer
(NSCLC), breast, ovarian, prostate, stomach, brain, colorectal cancer and melanoma
[9-16]. Furthermore, increased levels of EGFR gene expression is particularly linked
with higher cancer aggressiveness, tumor angiogenesis, postoperative adverse effects
as well as with chemotherapy and radiotherapy resistance. Therefore, all existing data
support the rational for the development of antibody-based or small molecule
inhibitors as EGFR-targeted therapeutics and presently, effective blockage of EGFR
2

ACCEPTED MANUSCRIPT
and HER-2 has been clinically validated in cancer therapy. Monoclonal antibodies
targeting the extracellular region of these receptors have been developed, with
trastuzumab (Herceptin®) being the first one approved [17-21]. In parallel, several
strategies involving small molecule tyrosine kinase inhibitors have been widely
investigated in recent years resulting in six inhibitors approved so far, mainly
targeting the EGFR (Fig. 1) [22-27]. All of the approved inhibitors act as ATP-
competitive inhibitors and bind to the ATP binding site of EGFR, effectively
preventing receptor phosphorylation and subsequent activation of downstream
pathways. They share a central 4-anilinoquinazoline core, which serves as hinge
adenine pocket binder, with the exception of osimertinib, which is structurally
different, presenting instead, a core anilinopyrimidine scaffold. The first EGFR-
targeted drugs are reversible inhibitors, offer great tolerability, and clinical outcome.
However, the emergence of acquired resistance to treatment by the first generation
inhibitors, prompted the discovery of next generation irreversible inhibitors such as
afatinib (Fig. 1). Afatinib, bears an electrophilic functional group, capable of covalent
Michael addition to a non- catalytic cysteine residue (Cys 797) of the enzyme [26].
Unfortunately, these drugs present a poor therapeutic window, as they are equally
active against wild type and mutant EGFR [28].
On the other hand, recently approved third generation drugs exemplified by
osimertinib (Fig. 1) are mutant selective irreversible inhibitors, capable of
overcoming the previously observed toxicity limitations [27]. They bind covalently to
Cys 797 and target different sensitizing EGFR mutant forms, including those
harboring the gatekeeper T790M point mutation [29]. Despite initial encouraging
results, additional resistance occurs rapidly, with one of the potential mechanisms
being the appearance of additional EGFR mutations such as C797S [11, 30].
Consequently, continuous research efforts in this field are ongoing, aiming to discover
new drug candidates in order to improve patient outcomes [31-34].
3

ACCEPTED MANUSCRIPT
Fig. 1. Approved EGFR inhibitors.
Among the approved drugs, lapatinib (Fig. 1) bears a suitably selected 4-(3-
fluorobenzyloxy)-3-chloroanilino-substitution in order to act as dual EGFR/HER2
inhibitor [35-37]. It was developed based on the consideration that dual inhibition
may lead to increased antitumor activity, since HER-2 overexpression potentiates
EGFR signaling [38]. Indeed, lapatinib is indicated for HER2 overexpressing
advanced or metastatic breast cancer [39, 40]. It is classified as I1/2 inhibitor, which
binds reversibly to the inactive conformation of the EGFR kinase, displacing the
characteristic aC-helix to the “out” conformation and giving the 3-fluorobenzyloxy
group access to the hydrophobic back pocket of the enzyme [41].
4

ACCEPTED MANUSCRIPT
Taking into account the intriguing features of lapatinib and related inhibitors
described in the literature [42-46], we have designed and synthesized a number of
modified purines falling under the general formula presented in Fig. 2. All of them
possess the hydrophobic 3-chloro-4-(3-fluorobenzyloxy)anilino-group, which
interacts with the selectivity pocket and extends into the back pocket of the enzyme,
considered an essential feature for dual EGFR and HER2 inhibition. The central
quinazoline core of lapatinib was replaced by purine isoster (Fig. 2, I-III) or purine
(Fig. 2, IV) hinge binders. The hydrophilic and solvent exposed 5-substituted furan
side chain of the drug was maintained, in order to optimize the drug-like properties of
the molecules. In parallel, the corresponding 5-furyl-unsubstituted derivatives were
synthesized, in order to evaluate the contribution of the furan side chain to the
potency of the compounds.
Fig. 2. General structure of the target compounds
2. Results and discussion
2.1. Theoretical studies
The design of the new analogues was based on the insight offered by the crystal
structure of lapatinib in complex with EGFR (pdb id.: 1XKK). The inhibitor binds the
inactive conformation of the kinase positioning the 3-fluorobenzyloxy side-chain in a
hydrophobic pocket formed next to the ATP binding pocket of the inactive-like kinase
by rearrangement of αC helix (Fig. 3A). In this case, lapatinib is mainly stabilized in
the kinase by a hydrogen bond formed at the enzyme hinge between the quinazoline
nitrogen N1 and the backbone amide of M793, whereas most additional interactions
5

ACCEPTED MANUSCRIPT
that contribute to the binding of the specific inhibitor are related to the substituents of
the quinazoline scaffold. The crystallographic data show that the 3-fluorobenzyl
group is critical for binding to the inactive kinase and hence for the inhibitor
selectivity, as it facilitates cation-π and extensive stacking interactions accommodated
within the additional pocket aside the binding site of the nucleotide. In a similar
fashion, the methanesulfonyl group is exposed to a considerable extent to the solvent,
further stabilizing the bound inhibitor by hydrogen bonds with polar sidechains of the
protein surface, including D800, R803 and R841. The detailed inspection of the
interaction mode between lapatinib and EGFR prompted the investigation of an
isosteric replacement of the quinazoline core by a ring system with the capacity to
accommodate additional hydrogen bonds with the kinase hinge. The imidazole-
containing fused scaffolds I-IV (Fig. 3B, 3C) were considered based on the above
strategy. Molecular simulations were performed and all four new scaffolds
representing stepwise isosteric replacements were flexibly docked to the EGFR
pocket and theoretically evaluated utilizing the Glide SP algorithm (Schrodinger Inc.)
[47-49]. The purine scaffold IV was expected to provide a model for directly
comparing the new isosteric heterocyclic system with lapatinib and assess whether an
additional hydrogen bonding site could potentially interact with the backbone
carbonyl of M793 in a bidentate fashion and further stabilize the inhibitor compared
to the original active lead. Furthermore, the extent by which either of the two
quinazoline nitrogens, N1 and N3, are critical for binding would be investigated from
the permutation by an aromatic carbon in scaffolds I-II, whereas the addition of a
fluorosubstituted carbon along the vector of the key hydrogen bonding interaction
stabilizing lapatinib in EGFR would offer structural justification as to whether
replacement of a hydrogen bond acceptor by a fluorine could have a favourable effect
to binding. Likewise, to probe for the importance of the methanesulfonyl substituent
to affinity, we prepared analogues that do not bear this group. Conversely, the anilino-
substituent of the drug was maintained, as the molecular modelling results
undoubtedly indicated its key importance for binding to the inactive state of EGFR,
thus providing selectivity over similar kinase targets.
6

ACCEPTED MANUSCRIPT
Fig. 3. A) A simplified depiction of the experimental binding mode of lapatinib in
EGFR. The inhibitor anchors into the enzyme pocket through a hydrogen bond
between the backbone of M793 and N1 of quinazoline whereas the 3-fluorobenzyloxy
side-chain is oriented toward an adjacent to the active site hydrophobic cavity and the
methanesulfonate group in the direction of bulk solvent. B) The designed isosteric
modification of quinazoline into a purine-like system carried by the novel analogues
aimed at achieving an additional hydrogen bond with respect to the original lead. C)
The four distinct heterocyclic ring systems considered and evaluated in this study with
the sites demonstrating hydrogen bond potential are depicted as red directed arrows.
2.2. Chemistry
In order to prepare compounds of the general structure I (Fig. 2), we used 4-amino-2-
chloropyridine (1, Scheme 1) as starting material, which was nitrated and the
resulting nitro-isomers 2 and 3 were chromatographically separated. Then, the 3-
nitroderivative 3 was reduced to give the diamine 4, which was ring-closed using 2-
furancarbaldehyde, to provide the substituted imidazopyridine 5 [50, 51]. This
compound was treated with 3-chloro-4-(3-fluorobenzyloxy)aniline (6), which was
7

ACCEPTED MANUSCRIPT
prepared according to a literature procedure, upon reaction of 2-chloro-4-nitrophenol
with 3-fluorobenzylchloride and reduction of the resulting nitroderivative [45].
However, the attempted nucleophilic substitution was not successful, despite the
variety of applied conditions (high temperature, microwave assistance, use of aniline
anion).
Cl
Cl
Cl
Cl
Cl
O2N
H2N
H2N
H2N
N
c
b
a
N
N
N
N
N
+
O
N
H
H2N
H2N
NO2
4
3
5
1
2
d
F
F
NH2
N
Cl
Cl
N
O
O
O
N
R
HN
HN
R
N
e
N
N
+
F
Cl
O
N
H2N
O
R
H
10: R=H
12: R=CH2NHCH2CH2SO2CH3
7: R=NO2
8: R=NH2
9: R=H
11: R=CH2NHCH2CH2SO2CH3
b
Scheme 1. Reagents and conditions. a. HNO3, H2SO4, 60oC, 4h; b. Zn, NH4Cl,
MeOH, 60oC, 2h; c. 2-furancarbaldehyde, DMSO, 100oC, 24h; d. 3-chloro-4-(3-
fluorobenzyloxy)aniline (6), Et3N, EtOH, reflux, 8h; e. 2-furancarbaldehyde or 17,
DMSO, 100oC, 3h.
Consequently, the synthetic procedure was modified and the substituted aniline 6 was
treated with the extremely reactive nitroderivative 3, thus providing the substituted
pyridine 7. Compound 7 was then reduced and the resulting aminoderivative 8 was
ring-closed using either 2-furancarbaldehyde, or freshly prepared suitably substituted
carbaldehyde 17 (Scheme 2). The latter was prepared using the commercially
available dioxolane 13, which was converted to the carbaldehyde 14 following a
literature procedure [52]. Compound 14 was then subjected to reductive amination
using 2-(methylsulfonyl)ethylamine [45] followed by addition of sodiumborohydride
8

ACCEPTED MANUSCRIPT
to give the furanacetal 16. Acidic hydrolysis of the acetal resulted in the
corresponding aldehyde 17.
Scheme 2. Reagents and conditions. a. (i) LDA in THF/n-heptane/ethylbenzene 2.0M,
DMF (dry), THF (dry), -78oC, 1h, (ii) rt, 24h; b. 2-(methylsulfonyl)ethylamine,
Na2SO4, MeOH, reflux, 3h; c. NaBH4, MeOH, rt, 2h; d. HCl 0.5Ν, CH2Cl2, rt, 1h.
The ring-closure of compound 8 (Scheme 1), provided the target imidazo[4,5-
c]pyridines 9 and 11, together with the isomeric 7-aminoimidazo[4,5-b]pyridines 10
and 12 respectively. Both isomers of each reaction were chromatographically
separated and characterized.
The conditions of the cyclization reaction were studied, concerning compounds 9 and
10, applying a range of temperature values and reaction times (Table 1). It was
assumed that the highest yield of the target product 9, as well as the optimum ratio of
9 over 10 were achieved when the reaction was carried out at 100oC for 3h.
Table 1. Study of the cyclization reaction conditions.
Temperature
Yield of 9
10 %
Yield of 10
62 %
Ratio 9 / 10
1 / 6
Reaction time
24 h
Rt
70 oC
33 %
17 %
2 / 1
24 h
9

ACCEPTED MANUSCRIPT
100 oC
130 oC
34 %
10 %
3 / 1
3 h
3 h
17 %
8 %
2 / 1
The second group of target compounds (compounds II, Fig. 2) was synthesized
according to the above described procedure, using 2-amino-4-chloropyridine (18,
Scheme 3) as starting material. This pyridine provided upon nitration both isomers 19
and 20 which were easily separated and identified [53]. Compounds 23 and 25,
together with the corresponding derivatives 24 and 26 were prepared accordingly
from the nitroderivative 20, and then separated and fully characterized.
Scheme 3. Reagents and conditions. a. HNO3, H2SO4, 60oC, 4h; b. 3-chloro-4-(3-
fluorobenzyloxy)aniline (6), Et3N, EtOH, reflux, 8h; c. Zn, NH4Cl, MeOH, 60oC, 2h;
d. 2-furancarbaldehyde or 17, DMSO, 100oC, 3h.
10

ACCEPTED MANUSCRIPT
The synthesis of the fluorosubstituted analogues (compounds III, Fig. 2) was
accomplished as depicted in Scheme 4, using 2-chloro-5-fluoropyridine (27). This
compound was converted to the corresponding N-oxide 28, which was successively
nitrated and reduced to give the aminopyridine 30. Following a second nitration step,
this pyridine was converted to the nitro-compound 31 [54] that served as key
intermediate for the preparation of the imidazopyridines 34 and 36, which were
obtained together with compounds 35 and 37 respectively.
Scheme 4. Reagents and conditions. a. H2O2 30%, TFA, 70oC, 4h; b. KNO3, H2SO4,
120oC, 2h; c. Fe, HCl, EtOH, reflux, 3h; d. KNO3, H2SO4, rt, 24h; e. 3-chloro-4-(3-
fluorobenzyloxy)aniline (6), Et3N, DMSO, 150oC, 24h; f. Zn, NH4Cl, MeOH, 60oC,
2h; g. 2-furancarbaldehyde or 17, DMSO, 100oC, 3h.
The synthesis of the purine analogues (compounds IV, Fig. 2) was performed
according to the procedure illustrated in Scheme 5, starting from the nitropyrimidine
38. This compound was treated in an one-pot reaction with the substituted aniline 6,
followed by the addition of ethanolic ammonia solution. This resulted in a mixture of
compound 39, together with an amount of the disubstituted derivative 41. The
chromatographic separation of these compounds proved to be very difficult, due to
11

ACCEPTED MANUSCRIPT
their extremely low solubility, thus, only a small amount of each derivative was
isolated in pure form using preparative thin layer chromatography, in order to
elucidate their structure. The mixture was introduced to the next, reduction step and
the aminoderivatives, 40 and 42, were isolated in pure form. Compound 40 was then
used for the preparation of the target derivatives 43 and 45, which resulted together
with 44 and 46, following the previously described methodology.
Scheme 5. Reagents and conditions. a. (i) 3-chloro-4-(3-fluorobenzyloxy)aniline (6),
NaHCO3, EtOH, 70oC, 1h (ii) NH3 in EtOH 4M, 70oC, 2h; b. Zn, NH4Cl, MeOH,
60oC, 2h; c. 2-furancarbaldehyde or 17, DMSO, 100oC, 3h.
2.3. Inhibition of Isolated Kinase and cell-based Phosphorylation Inhibition
The in vitro kinase inhibitory activity of the eight new compounds against EGFR was
evaluated using a biochemical assay measuring their ability to inhibit EGFR
autophosphorylation. The autophosphorylation inhibition potential of the derivatives
in comparison with lapatinib was investigated at two representative concentrations of
12

ACCEPTED MANUSCRIPT
10 µM and 1 µM (Fig. 4). Three compounds showed potent inhibitory activity against
EGFR (11, 23 and 45), with the effect being in direct comparison to corresponding
inhibition levels obtained by lapatinib.
Inhibitory Activity
100
90
80
70
60
50
40
30
20
10
0
Lapatinib
9
11
23
25
34
36
43
45
10 uM
1 uM
Fig. 4. EGFR-TK inhibitory activity of lapatinib and tested compounds derived by the
in vitro evaluation of the EGFR inhibitory activity in isolated enzyme. Compounds
were tested in duplicate and the reported % EGFR-TK inhibition represents the
average of three independent experiments (±SD).
Inhibition of phosphorylation of EGFR by the new compounds was further assessed in
the NSCLC cell line A549 and in the BrCa cell line HCC-1954, using a Western Blot
based assay (Fig. 5). Four derivatives (45, 25, 23 and 11) were found to possess
interesting inhibitory potencies, both in cell-free and cell-based experiments. The
compounds were also evaluated against the closely related HER2 kinase and 9
demonstrated the highest potency compared to lapatinib, followed by 43 (Table S1).
The ability of the compounds to inhibit HER2 phosphorylation in cells was also
investigated at two representative concentrations of 2 µM and 0.02 µM. Four
compounds (11, 25, 36 and 43) displayed the highest potencies compared to lapatinib
(Fig. S1).
13

ACCEPTED MANUSCRIPT
Fig. 5. In vitro inhibition of the phosphorylation of EGFR in A549 cells by Lapatinib
(1 µM & 0.01 µM) and the new analogues (10 µM & 1 µM).
2.4. Docking simulations
Flexible docking calculations using the Embrace module of Maestro were performed
to provide a possible explanation for the observed activity of the new derivatives [55].
With respect to the most active compounds 45 and 23, docking suggested that indeed
the N9 of novel cores II and IV demonstrates high predicted capacity to form an
additional hydrogen bond inside the kinase, in agreement with the rational design
discussed previously. This particular hydrogen bond is anticipated to further stabilize
the purine analogues in the enzyme pocket (Fig. 6) and possibly counterbalance
unfavorable entropic contributions originating from the restructuring of the water
network that might result by the isosteric replacement, thus preserving the original
kinase inhibitory potency of the quinazoline lead. Moreover, simulations provided a
clear explanation for the diminished inhibition of core III and the fluorosubstituted
derivatives 34 and 36, as this limited structural modification resulted in the disruption
of one out of the two hydrogen bonds stabilizing the ligand into the protein cavity.
Nevertheless, docking failed to fully explain the effect of the methanesulfonate
substituent of cores I and II to EGFR kinase binding affinity, with the results
indicating the chain bearing derivatives 11 and 25 as more active than their nor
counterparts 9 and 23, in contrast to the in vitro results suggesting that more than one
14

ACCEPTED MANUSCRIPT
binding geometries may be involved. Of note, the derived binding modes of the new
derivatives were regarded of high confidence, as their predicted binding geometries
demonstrated excellent agreement with the original lapatinib mode (Fig. 6).
Fig. 6. A superposition of the crystallographic binding mode of lapatinib (cyan
carbons, pbd id: 1XKK) and the predicted modes of active analogues 23 (green
carbons) and 45 (orange carbons) as derived by docking calculations inside the EGFR
kinase domain in a ribbon representation of the protein. In the inlet, a detailed
representation of the additional hydrogen bond hypothesized to further stabilize the
newly synthesized derivatives bearing an imidazopyridine (23) or purine scaffold (45)
at the kinase hinge, in comparison to the isosteric quinazoline core accommodating
only one favorable interaction inside the cavity of EGFR.
2.5. In vitro cytotoxic effect
The antiproliferative activity of the new compounds was investigated in the NSCLC
cell line A549 and in the BrCa cell line HCC-1954 using lapatinib as the reference
15

ACCEPTED MANUSCRIPT
compound (Fig. 7 a,b and Fig. S2). Four representative concentrations (60 µM, 40
µΜ, 10 µM and 1 µM) were used in order to identify the most potent compounds,
based on their overall cytotoxicity. Three compounds (43, 23 and 45) were found to
possess high cytotoxicities against both A549 and HCC-1954 cell lines (Fig. 7 a,b).
Overall, based on the in vitro evaluation results, derivatives 23 and 45 were selected
for further in vitro and pharmacokinetic evaluation in mice. These compounds were
potent inhibitors of EGFR phosphorylation (ELISA and Western blot assays) and at
the same time cytotoxic against both cell lines (A549 and HCC-1954). On the other
hand, compound 9 was found to display inhibitory effect in the biochemical tyrosine
kinase pHER2 assay, whereas 25 was found potent inhibitor in the Western Blot
Assay. However, neither of the two compounds showed strong antiproliferative
activity against the two cell lines (A549 and HCC-1954).
Fig. 7. Antiproliferative activity of lapatinib and selected potent new derivatives was
evaluated in (a) A549 and (b) HCC-1954 cells (60µM, 40 µΜ, 10 µM & 1 µM) by
MTT cytotoxicity assay.
The full antiproliferative activity profile of the two selected compounds (23 and 45)
was further evaluated in comparison to lapatinib in the two NSCLC cell lines A549
16

ACCEPTED MANUSCRIPT
and NCI-H1975 and the two BrCa cell lines HCC-1954 and MCF-7. Comparative
IC50 values are presented in Table 2.
Table 2. Antiproliferative effect of lapatinib, 23 and 45 against NSCLC cell lines
A549 and NCI-H1975 and BrCa cell lines HCC-1954 and MCF-7.
IC50 values (µM)*
Cell lines
Lapatinib
11 (±2)
9 (±1)
23
45
3 (±1)
33 (±10)
30 (±6)
14 (±6)
105 (±10)
A549
10 (±3)
9 (±3)
NCI-H1975
HCC-1954
MCF-7
5 (±1)
95 (±9)
126 (±13)
*Each IC50 value is the average of three independent experiments performed in
triplicate (±SD).
2.6. Characterization of selected compounds using Liquid Chromatography-Mass
Spectometry (LC-MS/MS) and Prediction of Key Efficacy Related Properties
LC-MS/MS characterization and method development-optimization was used for the
detection and quantification of lapatinib and the two selected compounds (23 and 45).
The identification of the characteristics of mass spec (MS) features (precursor and
product ions, table S2) helped us to proceed with the in vitro and in vivo studies that
were performed in order to assess the biological features of the molecules under
investigation (i.e. cellular uptake, in vivo biodistribution). Selective LC-MRM
(multiple reaction monitoring) methods were developed for the three compounds
(lapatinib, 23, 45) based on their characteristic MS features and the methodologies
were validated in the matrix of interest (i.e. cells collected from cell culture or mouse
blood) with respect to limit of quantification, linearity, and reproducibility. The
developed methodologies were utilized in cell uptake studies and pharmacokinetic
17

ACCEPTED MANUSCRIPT
evaluation of lapatinib and the two compounds. In addition, in silico predictions were
made for some key molecular properties concerning the efficacy of these compounds
(Table S2). Among them, 23 displayed higher cLogP (4.7) than lapatinib and 45 (3.7
and 2.9 respectively) therefore being the most lipophilic, while its polar surface area
was 67.2 Å2 (<100 Å2), a desired feature for drug candidates, known to influence cell
membrane permeability [56].
2.7. Cell uptake of 23 and 45 in A549 cells
The cell uptake of 23 and 45 in A549 cells was evaluated in comparison to lapatinib
(Fig. 8). Compound 23 exhibited higher cellular uptake in A549 cells than lapatinib,
especially with respect to the late time points of 24h and 48h of incubation time
(AUC(1-48h): 1.5 nmol/100.000 cells x h versus AUC(1-48h) :1.06 nmol/100.000 cells x h
for lapatinib). It could be suggested that the physiochemical properties of this
compound (higher lipophilicity, lower total polar surface area) are responsible for its
higher accumulation [57].
Fig. 8. Cellular uptake of 23 and 45 in the NSCLC cell line A549, in comparison to
Lapatinib. Cells were incubated with 1 µM Lapatinib, 23 and 45 for selected time
points (1h, 4h, 8h, 24h and 48h) and were then lysed in order to determine
18

ACCEPTED MANUSCRIPT
intracellular levels of the three compounds by LC-MS/MS. Experiments were
performed in triplicate (±SD)
2.8. Pharmacokinetic evaluation of 23 and 45
Assessment of the in vivo pharmacokinetics (PK) of the two selected compounds (23
and 45) was performed in mice. The PK parameters of both molecules were compared
to lapatinib after oral administration, using an identical solubilizing vehicle and an
equimolar dose of 34 µmol/kg (Fig. 9). Lapatinib maximum concentration in blood
was achieved at 2h post administration (1.47 µM) with an AUC(0-24h) of 6.7 x h x µΜ.
Similarly, 23 and 45 reached highest blood concentrations at 2h post administration
(0.30 µΜ and 0.07 µΜ for 45). Overall, both compounds had no advantages over
lapatinib in oral dosing. However, the pharmacokinetic experiment was used as a
guide for subsequent studies for compound 23. From the two compounds, 23 had only
marginally improved bioavailability compared to 45 (AUC(0-24h) of 2.1 h x µΜ for 23
compared to 0.87 x h x µΜ for 45). Based on the above mentioned data, 23 was
selected as lead candidate for subsequent efficacy studies in mice.
19

ACCEPTED MANUSCRIPT
Fig. 9. Pharmacokinetic evaluation of 23 and 45 versus Lapatinib. Male C57BL/6N
mice (n = 5) were dosed orally with either Lapatinib, 23 or 45 at a dose of 34 µmol/kg
and blood samples were collected at selected time points. Lapatinib, 23 and 45 levels
were monitored by LC-MS/MS. The areas under the curve (AUCs) for each treatment
were calculated as a measure of drug exposure over time (± SD).
2.9. In vivo antitumor efficacy of 23 and Lapatinib
The efficacy of 23 was evaluated on a EGFR positive xenograft lung cancer model
using the NSCLC A549 cell line. Efficacy was compared to equimolar dose of
lapatinib under a selected dosing scheme (86 µmol/kg, once daily oral
administration). Pharmacological treatment with either lapatinib or 23 was initiated
when tumors reached approximately 150 mm3 and was continued for a total period of
27 days. At the end of the treatment period (day 27), the average tumor volume in
mice treated with 23 (1311 ± 458 mm3, P= 0.01) or with lapatinib (1338 ± 358 mm3,
P= 0.001) was significantly lower compared to the size of tumors for vehicle (1869 ±
134 mm3) mice (Fig. 10a) suggesting that the two compounds (lapatinib and 23) have
similar efficacy profiles in the specific animal model. No signs of discomfort or
evidence of drug-induced toxicity was observed as measured by changes in animal
total body weight (Fig. 10b). Concentrations of lapatinib and 23 in blood and tumor
tissue were measured at 2 h after the final dose. Lapatinib’s blood concentrations were
2.81 µΜ, whereas the blood levels of 23 were 0.64 µM (Fig. 10c). These results are
consistent with the pharmacokinetic evaluation of the two compounds, where an
equimolar dose of 34 µmol/kg was used. Although, we were able to measure lapatinib
levels in tumor tissue, levels of compound 23 had great variation, possibly due to a
different tumor uptake profile due to its different physiochemical properties that could
lead to differences in the tumor uptake mechanisms (e.g. P-gp efflux, passive
diffusion) or the formation of metabolites that couldn’t be identified.
20

ACCEPTED MANUSCRIPT
Figure 10. Therapeutic efficacy of 23. (a) Tumor growth inhibition in NOD/SCID mice xenografted with
A549 cells and treated with daily gavage administration of 23, lapatinib and vehicle. Each point represents the
mean of at least 8 tumor volumes resulting from at least 4 mice ± SD ***, P < 0.001 vs controls and ± SD **,
P < 0.01 vs controls. (b) Absence of toxicity was assessed by total body weight measurements of all treated
mice. (c) Average blood drug concentrations were measured by LC-MS/MS at 2h post a final dose on d27, (±
SD).
Overall, this study deals with the synthesis, biological and pharmacokinetic evaluation
of a number of new lapatinib analogues, designed to assess the potential of applying
rationally selected isosteric replacements on the lead quinazoline core of the drug.
The results could assist in expanding the SAR landscape of EGFR inhibition and,
moreover probe for the importance of specific structural features on the activity
against the target kinase. The in vitro evaluation of the synthesized compounds in a
cell free assay showed that derivative 45, possessing a purine central core instead of
the quinazoline ring of lapatinib, showed a highly promising efficacy, which was
perfectly anticipated by docking simulations. Replacement of the quinazoline N1 by a
fluorosubstituted carbon perturbed the pharmacophoric hydrogen bond acceptor and
thus resulted in inactive analogues (34, 36). However, a consistent trend with respect
to the presence of the furan side chain on inhibition potency could not be established,
as 23 and 11 were determined to be equally active against EGFR kinase, regardless of
their difference with respect to the aforementioned substitution. In addition, the
corresponding derivatives 25 (carrying substitution on the furan ring) and 9 (without
substitution) both showed diminished potency.
21

ACCEPTED MANUSCRIPT
To further confirm the biological activities obtained in the isolated enzyme, the
inhibitory activity of the novel derivatives against EGFR was carried out in a cell-
based assay. The observed inhibitory effect of the new analogues on EGFR
autophosphorylation was in good agreement with the results obtained at the cell free
assay as 11, 23 and 45 were the most effective in inhibiting phospho-EGFR
formation, with the exception of derivative 25. In order to study the effect that the
new analogues could exert on cell proliferation, the same cells were treated with the
synthesized compounds and their cytotoxic effect was measured, showing that 23, 43
and 45 were the most potent in this respect. Whereas two of the compounds, 23 and
45, were well expected to exert potent cytotoxic results based on their already
confirmed EGFR inhibition potency, the discrepancy between the EGFR inhibitory
potential and cytotoxic potency of 43 was an indication of considerable off-target
activities, possibly leading to the observed cytotoxicity. In light of those observed
biological responses and aiming to assist the evaluation, the antiproliferative effect of
the most promising derivatives 23 and 45 was studied in a more extensive protocol
using relevant cell lines, in order to obtain a more detailed description of the in vitro
cytotoxic landscape of the studied molecules. Compound 23 exhibited higher potency
than 45 in both NSCLC cell lines, as well as against BrCa cell line HCC-1954, and
most importantly it displayed greater potency (IC50 = 3 µM) compared to lapatinib
(IC50 = 11 µM) in the A549 cell line. In line with lapatinib, those two compounds
could not inhibit strongly MCF-7 cell proliferation, presenting comparable IC50
values, approximately equal to 100 µM. As this breast cancer cell line does not
express EGFR, the finding suggests that both compounds could be regarded as
considerably selective for EGFR. 24 and 45 were then subsequently assessed for their
bioavailability and cellular uptake and 23 resulted in high measured levels, both in
cells and blood, suggesting that 23 could lead to more sustainable intracellular levels,
compared to lapatinib. Importantly, the improved pharmacokinetic profile of 23 was
in line with the more relevant drug-like physicochemical properties such as cLogP of
the specific substance. Thus, 23 could be perceived as a valuable drug candidate, with
enhanced tumor bioavailability, more importantly considering the comparable in vivo
efficacy to lapatinib.
3. Conclusion
22

ACCEPTED MANUSCRIPT
In conclusion, we have designed and synthesized a new series of lapatinib analogues,
inducing modifications in the central quinazoline core and taking care to preserve, or
not extensively modify, the original substituents of the drug. The chemical procedure
for the preparation of the compounds involved the use of a suitable
aminonitrochloropyridine (or pyrimidine), which was first substituted with a
preformed aniline and then reduced and ring-closed upon reaction with selected
furancarbaldehydes. This procedure resulted in the preparation of a series of
corresponding isomeric side-products as well, bearing a primary amino-group, thus
considered as adenine isosters. These compounds were also introduced in the primary
screening tests, albeit they did not possess interesting inhibitory activity against
EGFR in the biochemical assays and thus, were not further investigated. Following
the evaluation of the biochemical and cellular activity of the target compounds against
EGFR and HER2, two derivatives, 23 and 45 were selected for further in vitro and in
vivo evaluation. We have shown that 23 had greater potency than lapatinib against the
A549 cells based on the cell toxicity studies. Importantly, 23 presented satisfactory
pharmacokinetic properties and higher intracellular levels than lapatinib in the A549
cells, possibly due to its higher lipophilicity and enhanced tumor penetration. The in
vivo efficacy study of 23 in an EGFR xenograft animal model showed that it can
successfully inhibit tumor growth, with a similar efficacy profile with that of lapatinib
and with minimal phenotypic toxicity.
In conclusion, compound 23 possessing an altered, nitrogen rich central scaffold
compared to lapatinib, is a potent and promising novel anticancer compound with
adequate physiochemical properties and interesting biological activity in EGFR
inhibition.
4. Experimental Section
4.1. Chemistry
1
Melting points were determined on a Büchi apparatus and are uncorrected. H NMR
spectra and 2D spectra were recorded on a Bruker Avance III 600 or a Bruker Avance
13
DRX 400 instrument, whereas C NMR spectra were recorded on a Bruker Avance
III 600 or a Bruker AC 200 spectrometer in deuterated solvents and were referenced
23

ACCEPTED MANUSCRIPT
to TMS (δ scale). The signals of 1H and 13C spectra were unambiguously assigned by
using 2D NMR techniques: 1H1H COSY, NOESY, HMQC, and HMBC. Mass spectra
were recorded with a LTQ Orbitrap Discovery instrument, possessing an Ionmax
ionization source. Flash chromatography was performed on Merck silica gel 60 (0.040
- 0.063 mm). Analytical thin layer chromatography (TLC) was carried out on
precoated (0.25 mm) Merck silica gel F-254 plates.
4.1.1. 4-Chloro-2-(furan-2-yl)-1H-imidazo[4,5-c]pyridine (5). To a solution of
compound 4 (300 mg, 2.09 mmol)[51] in dimethylsulfoxide (1.0 mL) was added
under Ar a solution of 2-furaldehyde in dimethylsulfoxide (0.5 mL, 2.09 mmol) and
o
the mixture was heated at 100 C for 24 h. The solvent was then vacuum-evaporated
and the residue was purified using silica gel column chromatography
(cyclohexane/ethyl acetate: 1/2), in order to provide 5 as a light brown solid (240 mg,
52%). M.p. > 215 oC dec. (Et2O). 1H NMR (600 MHz, DMSO-d6) δ 13.95-13.35 (brs,
1H, NH), 8.11 (d, J = 5.5 Hz, 1H, H-6), 8.03 (dd, J = 1.8, 0.6 Hz, 1H, H-5’), 7.55 (d, J
= 5.5 Hz, 1H, H-7), 7.39 (d, J = 3.4 Hz, 1H, H-3’), 6.78 (dd, J = 3.4, 1.8 Hz, 1H, H-
4’). 13C NMR (151 MHz, DMSO-d6) δ 145.90 (C-5’), 144.27 (C-2’), 141.99 (C-2),
141.60 (C-4), 141.01 (C-6), 139.44 (C-7a), 136.98 (C-3a), 112.92 (C-3’), 112.70 (C-
4’), 107.95 (C-7).
4.1.2. 2-N-[3-Chloro-4-(3-fluorobenzyloxy)phenyl]-3-nitropyridine-2,4-diamine (7).
To a solution of the pyridine 3 (300 mg, 1.73 mmol)[51] in absolute ethanol (10 mL)
were added triethylamine (0.5 mL, 3.58 mmol) and the aniline 6 (500 mg, 1.99
mmol)[45] and the mixture was heated at reflux for 8 h. The solvent was vacuum-
evaporated, water was added to the residue and the precipitate was filtered, washed
with water, dried and purified using silica gel column chromatography
o
(dichloromethane) to give pure 7 (440 mg, 65%). M.p. = 159-160 C (EtOAc/n-
hexane). 1H NMR (400 MHz, DMSO-d6) δ 10.39 (s, 1H, NH), 8.20 (s, 2H, NH2), 7.91
(d, J = 2.4 Hz, 1H, H-2’), 7.69 (d, J = 5.9 Hz, 1H, H-6), 7.46 (m, 2H, H-6’, H-5’’),
7.30 (m, 2H, H-2’’, H-6’’), 7.17 (m, 2H, H-5’, H-4’’), 6.30 (d, J = 5.9 Hz, 1H, H-5),
5.22 (s, 2H, CH2). 13C NMR (151 MHz, DMSO-d6) δ 162.99, 161.37 (C-3’’), 153.04
(C-2), 151.53 (C-4), 149.88 (C-6), 149.58 (C-4’), 139.71, 139.66 (C-1’’), 133.10 (C-
24

ACCEPTED MANUSCRIPT
1’), 130.55, 130.49 (C-5’’), 124.30 (C-2’), 123.26 (C-6’’), 122.50 (C-6’), 121.05 (C-
3’), 116.50 (C-3), 114.71, 114.57 (C-4’’), 114.27 (C-5’), 114.03, 113.89 (C-2’’),
103.77 (C-5), 69.35 (CH2).
4.1.3. 2-N-[3-Chloro-4-(3-fluorobenzyloxy)phenyl]pyridine-2,3,4-triamine (8). To a
solution of the nitroderivative 7 (900 mg, 2.32 mmol) in methanol (10 mL) were
added zinc dust (2.60 g, 39.8 mmol) and ammonium chloride (2.90 g, 54.2 mmol) and
o
the mixture was heated at 60 C for 2 h. It was then filtered through a celite pad and
the filtrate was concentrated under vacuum. The residue was purified by silica gel
column chromatography (CH2Cl2/MeOH: 85/15) to give 8 as an oil (600 mg, Yield:
1
72%). H NMR (600 MHz, DMSO-d6) δ 7.76 (d, J = 2.6 Hz, 1H, H-2’), 7.59 (s, 1H,
NH), 7.44 (m, 1H, H-5’’), 7.33 (dd, J = 9.0, 2.6 Hz, 1H, H-6’), 7.31-7.26 (m, 3H, H-6,
H-2’’, H-6’’), 7.15 (m, 1H, H-4’’), 7.07 (d, J = 9.0 Hz, 1H, H-5’), 6.20 (d, J = 5.4 Hz,
1H, H-5), 5.36 (s, 2H, 4-NH2), 5.14 (s, 2H, CH2), 4.13 (s, 2H, 3-NH2). 13C NMR (151
MHz, DMSO-d6) δ 162.99, 161.38 (C-3’’), 146.47 (C-4’), 143.87 (C-2), 142.39 (C-4),
140.09, 140.04 (C-1’’), 137.83 (C-1’), 136.22 (C-6), 130.47, 130.42 (C-5’’), 123.28
(C-6’’), 121.44 (C-3’), 119.01 (C-2’), 117.34 (C-6’), 115.24 (C-5’), 115.17 (C-3),
114.60, 114.47 (C-4’’), 114.04, 113.89 (C-2’’), 104.58 (C-5), 69.74 (CH2).
4.1.4.
N-[3-Chloro-4-(3-fluorobenzyloxy)phenyl]-2-(furan-2-yl)-1H-imidazo[4,5-
c]pyridin-4-amine (9) and 3-[3-chloro-4-(3-fluorobenzyloxy)phenyl]-2-(furan-2-yl)-
3H-imidazo[4,5-b]pyridin-7-amine (10). To a solution of 8 (120 mg, 0.33 mmol) in
dimethylsulfoxide (1.0 mL) was added under Ar a solution of 2-furaldehyde (28 µL,
0.34 mmol) in dimethylsulfoxide (0.1 mL) and the resulting solution was heated at
o
100 C in a preheated oilbath for 3 h. The solvent was vacuum-evaporated and the
residue was purified using silica gel column chromatography (cyclohexane/EtOAc:
1/2), to provide both isomers, which were separated using preparative thin layer
chromatography (cyclohexane/EtOAc: 2/3).
Data for 9: 50 mg, 34%, M.p. = 110-111 oC (MeOH). 1H NMR (400 MHz, DMSO-d6)
δ 13.21 (s, 1H, imidazole NH), 9.13 (s, 1H, aniline NH), 8.33 (d, J = 2.5 Hz, 1H, H-
2’), 7.96 (d, J = 1.3 Hz, 1H, H-5’’’), 7.87 (m, 2H, H-6, H-6’), 7.45 (m, 1H, H-5’’),
25

ACCEPTED MANUSCRIPT
7.31 (m, 2H, H-2’’, H-6’’), 7.20 (d, J = 3.4 Hz, 1H, H-3’’’), 7.15 (m, 2H, H-5’, H-
4’’), 6.96 (d, J = 5.7 Hz, 1H, H-7), 6.75 (dd, J = 3.4, 1.3 Hz, 1H, H-4’’’), 5.19 (s, 2H,
CH2). 13C NMR (50 MHz, DMSO-d6) δ 164.62, 159.78 (C-3’’), 147.58 (C-4), 147.39
(C-4’), 145.08 (C-2’’’), 144.72 (C-5’’’), 142.11 (C-2), 140.08, 139.93 (C-1’’), 139.75
(C-6), 138.65 (C-7a), 136.06 (C-1’), 130.58, 130.41 (C-5’’), 128.14 (C-3a), 123.28
(C-6’’), 121.14 (C-3’), 120.34 (C-2’), 118.71 (C-6’), 114.79 (C-5’), 114.79, 114.37
(C-4’’), 114.20, 113.77 (C-2’’), 112.36 (C-4’’’), 110.62 (C-3’’’), 99.77 (C-7), 69.56
(CH2). HR-MS (ESI) m/z: calcd for C23H17ClFN4O2, [M+H]+ = 435.1019, found
435.1010.
Data for 10: 15 mg, 10%, M.p. = 200-201 oC (EtOAc/n-hexane). 1H NMR (400 MHz,
DMSO-d6) δ 7.80 (s, 1H, H-5’’’), 7.74 (d, J = 5.4 Hz, 1H, H-5), 7.67 (s, 1H, H-2’),
7.50 (m, 1H, H-5’’), 7.41 (m, 2H, H-5’, H-6’), 7.36 (m, 2H, H-2’’, H-6’’), 7.21 (m,
1H, H-4’’), 6.58 (brs, 2H, NH2), 6.55 (s, 1H, H-4’’’), 6.42 (d, J = 5.4 Hz, 1H, H-6),
6.20 (d, J = 3.1 Hz, 1H, H-3’’’), 5.36 (s, 2H, CH2).13C NMR (151 MHz, DMSO-d6) δ
163.03, 161.42 (C-3’’), 153.79 (C-4’), 149.46 (C-3a), 146.90 (C-7), 145.16 (C-5),
144.55 (C-5’’’), 144.26 (C-2’’’), 140.20 (C-2), 139.26, 139.21 (C-1’’), 130.67, 130.62
(C-5’’), 129.90 (C-2’), 128.82 (C-1’), 128.47 (C-6’), 123.42 (C-6’’), 122.33 (C-7a),
121.79 (C-3’), 114.92, 114.78 (C-4’’), 114.46 (C-5’), 114.21, 114.07 (C-2’’), 111.78
(C-4’’’), 111.64 (C-3’’’), 103.09 (C-6), 69.52 (CH2). HR-MS (ESI) m/z: calcd for
C23H15ClFN4O2, [M-H]- = 433.0862, found 433.0876.
4.1.5. 2-{5-[2-(Methylsulfonyl)ethylamino]methyl)furan-2-yl}-1,3-dioxolane (16). To
a solution of carbaldehyde 14 (310 mg, 1.85 mmol)[52] in dry methanol (7.0 mL)
were added under Ar sodium sulfate (1.20 g, 8.45 mmol) followed by 2-
(methylsulfonyl)ethylamine (310 mg, 2.52 mmol) and the mixture was heated at
reflux for 3 h. Then, the reaction was allowed to return at rt, sodium borohydride (370
mg, 9.78 mmol) was added under cooling and stirring continued at rt for 2 h. The
solvent was vacuum-evaporated, water was added to the residue and it was extracted
with dichloromethane. The organic layers were dried (sodium sulfate) and evaporated
1
to dryness to give pure 16 as a yellow oil (500 mg, 98%). H NMR (600 MHz,
DMSO-d6) δ 6.44 (d, J = 3.1 Hz, 1H, H-4), 6.24 (d, J = 3.1 Hz, 1H, H-3), 5.81 (s, 1H,
H-2’), 4.03 – 3.88 (m, 4H, H-4’, H-5’), 3.68 (s, 2H, CH2(a)), 3.21 (t, J = 6.6 Hz, 2H,
26

ACCEPTED MANUSCRIPT
CH2(c)), 3.00 (s, 3H, CH3), 2.91 (t, J = 6.6 Hz, 2H, CH2(b)), 2.40 – 2.30 (brs, 1H,
13
NH). C NMR (151 MHz, DMSO-d6) δ 154.66 (C-2), 149.87 (C-5), 109.58 (C-4),
107.28 (C-3), 96.80 (C-2’), 64.46 (C-4’, C-5’), 53.60 (CH2(c)), 44.95 (CH2(a)), 42.05
(CH2(b)), 41.45 (CH3).
4.1.6. 5-{[2-(Methylsulfonyl)ethylamino]methyl}furan-2-carboxaldehyde (17). A 0.5N
HCl solution (5.0 mL, 2.5 mmol) was added dropwise under cooling to a solution of
the acetal 16 (220 mg, 0.80 mmol) in dichloromethane (10 mL) and the mixture was
stirred vigorously at rt for 1 h. Then, a 25% ammonium hydroxide solution was added
until pH 11-12. Upon extraction with dichloromethane, the organic layer was dried
(sodium sulfate) and evaporated to dryness to provide pure 17 as a yellow oil (170
mg, 93%). 1H NMR (600 MHz, DMSO-d6) δ 9.52 (s, 1H, CHO), 7.49 (d, J = 3.5 Hz,
1H, H-3), 6.60 (d, J = 3.5 Hz, 1H, H-4), 3.82 (s, 2H, CH2(a)), 3.24 (t, J = 6.7 Hz, 2H,
13
CH2(c)), 3.00 (s, 3H, CH3), 2.93 (t, J = 6.7 Hz, 2H, CH2(b)). C NMR (151 MHz,
DMSO-d6) δ 177.72 (CHO), 161.34 (C-5), 151.75 (C-2), 124.61 (C-3), 110.05 (C-4),
53.59 (CH2(c)), 45.02 (CH2(a)), 42.12 (CH2(b)), 41.51 (CH3).
4.1.7.
(methylsulfonyl)ethylaminomethyl]furan-2-yl}-1H-imidazo[4,5-c]pyridin-4-amine
(11) and 3-[3-chloro-4-(3-fluorobenzyloxy)phenyl]-2-{5-[2-
N-[3-Chloro-4-(3-fluorobenzyloxy)phenyl]-2-{5-[2-
(methylsulfonyl)ethylaminomethyl]furan-2-yl}-3H-imidazo[4,5-b]pyridin-7-amine
(12). These compounds were prepared using a methodology analogous to that of 9 and
10 respectively and purified first by silica gel column chromatography
(EtOAc/MeOH: 8/2) resulting to a mixture of both products, which were then isolated
pure using semi-preparative reverse phase high performance liquid chromatography
(H2O/MeOH: 2/8 to 1/9).
o
1
Data for 11: 30 mg, Yield: 20%, M.p. = 101-102 C (MeOH). H NMR (600 MHz,
DMSO-d6) δ 13.56 – 12.56 (brs, 1H, imidazole NH), 8.97 (s, 1H, aniline NH), 8.33
(d, J = 2.5 Hz, 1H, H-2’), 7.82 (m, 2H, H-6, H-6’), 7.46 (m, 1H, H-5’’), 7.31 (m, 2H,
H-2’’, H-6’’), 7.19 – 7.06 (m, 3H, H-5’, H-4’’, H-3’’’), 6.94 (d, J = 5.6 Hz, 1H, H-7),
6.51 (d, J = 3.1 Hz, 1H, H-4’’’), 5.19 (s, 2H, OCH2), 3.83 (s, 2H, CH2(a)), 3.27 (t, J =
27

ACCEPTED MANUSCRIPT
6.8 Hz, 2H, CH2(c)), 3.02 (s, 3H, CH3), 2.98 (t, J = 6.8 Hz, 2H, CH2(b)), 2.48 – 2.40
(brs, 1H, ethylamine NH). 13C NMR (151 MHz, DMSO-d6) δ 163.04, 161.43 (C-3’’),
156.14 (C-5’’’), 147.51 (C-4), 147.41 (C-4’), 144.10 (C-2’’’), 142.25 (C-2), 140.05,
140.01 (C-1’’), 139.72 (C-6), 138.70 (C-7a), 136.11 (C-1’), 130.57, 130.52 (C-5’’),
128.22 (C-3a), 123.35 (C-6’’), 121.21 (C-3’), 120.33 (C-2’), 118.70 (C-6’), 114.88
(C-5’), 114.70, 114.56 (C-4’’), 114.09, 113.95 (C-2’’), 111.49 (C-3’’’), 109.56 (C-
4’’’), 99.83 (C-7), 69.60 (OCH2), 53.67 (CH2(c)), 45.14 (CH2(a)), 42.12 (CH2(b)),
41.49 (CH3). HR-MS (ESI) m/z: calcd for C27H24ClFN5O4S, [M-H]- = 568.1216,
found 568.1193.
Data for 12: 10 mg. Oil. Yield: 7%. 1H NMR (600 MHz, DMSO-d6) δ 7.73 (d, J = 5.5
Hz, 1H, H-5), 7.66 (m, 1H, H-2’), 7.50 (m, 1H, H-5’’), 7.40 (m, 2H, H-5’, H-6’), 7.35
(m, 2H, H-2’’, H-6’’), 7.21 (m, 1H, H-4’’), 6.53 (brs, 2H, NH2), 6.41 (d, J = 5.5 Hz,
1H, H-6), 6.33 (d, J = 3.4 Hz, 1H, H-4’’’), 6.00 (d, J = 3.4 Hz, 1H, H-3’’’), 5.37 (s,
2H, OCH2), 3.69 (s, 2H, CH2(a)), 3.21 (t, J = 6.6 Hz, 2H, CH2(c)), 3.00 (s, 3H, CH3),
2.90 (t, J = 6.6 Hz, 2H, CH2(b)), 2.35 – 2.29 (brs, 1H, NH). 13C NMR (151 MHz,
DMSO-d6) δ 163.07, 161.46 (C-3’’), 155.92 (C-5’’’), 153.85 (C-4’), 149.73 (C-7),
146.74 (C-3a), 145.27 (C-5), 143.21 (C-2’’’), 140.29 (C-2), 139.29, 139.24 (C-1’’),
130.73, 130.67 (C-5’’), 130.05 (C-2’), 128.91 (C-1’), 128.63 (C-6’), 123.47 (C-6’’),
122.34 (C-7a), 121.87 (C-3’), 114.97, 114.83 (C-4’’), 114.57 (C-5’), 114.25, 114.10
(C-2’’), 112.35 (C-3’’’), 108.92 (C-4’’’), 103.14 (C-6), 69.56 (OCH2), 53.65
(CH2(c)), 44.93 (CH2(a)), 42.08 (CH2(b)), 41.55 (CH3). HR-MS (ESI) m/z: calcd for
C27H26ClFN5O4S, [M+H]+ = 570.1373, found 570.1346.
4.1.8. 4-N-[3-Chloro-4-(3-fluorobenzyloxy)phenyl]-3-nitropyridine-2,4-diamine (21).
This compound was prepared using a methodology analogous to that of 7, starting
from the aminopyridine 20 [53]. The compound was purified by silica gel column
o
chromatography (CH2Cl2/MeOH: 95/5). Yield: 89%. M.p. > 210 C dec. (EtOAc/n-
hexane). 1H NMR (600 MHz, DMSO-d6) δ 10.06 (s, 1H, NH), 7.86 (s, 2H, NH2), 7.69
(d, J = 5.9 Hz, 1H, H-6), 7.50 – 7.45 (m, 2H, H-2’, H-5’’), 7.34 – 7.25 (m, 4H, H-5’,
H-6’, H-2’’, H-6’’), 7.18 (m, 1H, H-4’’), 5.91 (d, J = 5.9 Hz, 1H, H-5), 5.27 (s, 2H,
CH2). 13C NMR (151 MHz, DMSO-d6) δ 163.02, 161.40 (C-3’’), 156.16 (C-2),
153.00 (C-6), 151.89 (C-4’), 150.51 (C-4), 139.44, 139.39 (C-1’’), 131.78 (C-1’),
28