Investigating molecular dynamics-guided lead optimization of EGFR inhibitors

Accepted Manuscript

Martin J. Lavecchia, Raimon Puig de la Bellacasa, José I. Borrell, Claudio N.

Cavasotto

PII:

S0968-0896(15)30211-X

http://dx.doi.org/10.1016/j.bmc.2015.12.046

BMC 12739

DOI:

Reference:

Bioorganic & Medicinal Chemistry

To appear in:

Received Date:

Revised Date:

Accepted Date:

28 October 2015

18 December 2015

28 December 2015

Please cite this article as: Lavecchia, M.J., Bellacasa, R.P.d., Borrell, J.I., Cavasotto, C.N., Investigating molecular

dynamics-guided lead optimization of EGFR inhibitors, Bioorganic & Medicinal Chemistry (2015), doi: http://

dx.doi.org/10.1016/j.bmc.2015.12.046

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.

Investigating molecular dynamics-guided lead optimization of

EGFR inhibitors

Martin J. Lavecchia^1,#, Raimon Puig de la Bellacasa^2,#, José I. Borrell², Claudio N.

Cavasotto^1,*

¹Instituto de Investigación en Biomedicina de Buenos Aires (IBioBA) - CONICET - Partner

Institute of the Max Planck Society, Godoy Cruz 2390, C1425FQD, Buenos Aires,

Argentina.

²Grup d’Enginyeria Molecular (GEM), IQS School of Engineering, Universitat Ramon Llull,

Via Augusta 390 Barcelona E-08017, Spain.

^*Corresponding author: Claudio N. Cavasotto, Tel: +54 11 4899-5500 e-mail cnc@cavasotto-

lab.net, ccavasotto@ibioba-mpsp-conicet.gov.ar

# Equally contributed to this work.

E-mail addresses: lavecchia@gmail.com mailto:(M.J. Lavecchia); raimon.puig@iqs.url.edu

(R. Puig de la Bellacasa); j.i.borrell@iqs.url.edu (J.I. Borrell); cnc@cavasotto-lab.net,

ccavasotto@ibioba-mpsp-conicet.gov.ar (C.N. Cavasotto).

1

Graphical abstract

Keywords

EGFR; molecular docking; molecular dynamics; ligand binding free energy calculation;

MM/GBSA; small-molecule inhibitor; pyrido[2,3-d]pyrimidine; hit-to-lead optimization.

Abbreviations

EGFR, Epidermal Growth Factor Receptor; MD, molecular dynamics; MM/GBSA,

Molecular Mechanics/Generalized Born Surface Area; MM/PBSA, Molecular

Mechanics/Poisson-Boltzmann Surface Area; TK, Tyrosine Kinase; Generalized Amber Force

Field (GAFF); Solvent-accessible Surface Area (SASA); TKI, Tyrosine Kinase Inhibitor;

PDB, Protein Data Bank.

.

Abstract

The Epidermal Growth Factor Receptor (EGFR) is part of an extended family of proteins that

together control aspects of cell growth and development, and thus a validated target for drug

discovery. We explore in this work the suitability of a molecular dynamics-based end-point

binding free energy protocol to estimate the relative affinities of a virtual combinatorial

library designed around the EGFR model inhibitor 6{1} as a tool to guide chemical synthesis

toward the most promising compounds. To investigate the validity of this approach, selected

analogues including some with better and worse predicted affinities relative to 6{1} were

synthesized, and their biological activity determined. To understand the binding determinants

of the different analogues, hydrogen bonding and van der Waals contributions, and water

molecule bridging in the EGFR-analogue complexes were analyzed. The experimental

validation was in good qualitative agreement with our theoretical calculations, while also a 6-

dibromophenyl-substituted compound with enhanced inhibitory effect on EGFR compared to

the reference ligand was obtained.

2

1. Introduction

The epidermal growth factor receptor (EGFR) is one of the members of the family of

tyrosine kinases (TKs) that are involved in the modulation of growth factor signaling.

Receptors in this family contain an extracellular ligand binding domain, a transmembrane

region, and an intracellular tyrosine kinase domain. Upon binding to a growth factor,

receptors in this family dimerize, thus activating their kinase domain, and triggering

intracellular signaling pathways, which control tumor cell growth, proliferation, survival,

metastasis and angiogenesis¹.

Mutations that lead to EGFR overexpression (known as upregulation) have been

3

associated with

a

number of cancers, including mammary^2,

,

ovarian⁴,

esophageal⁵, squamous cell head and neck carcinomas⁶, non-small cell lung cancer⁷,

glioblastoma⁸, where it correlates with poor prognosis⁹. Consequently, a huge effort has been

poured in the development of anticancer drugs directed targeting EGFR¹⁰which include, on

one side, monoclonal antibodies targeting the extracellular domain, such as cetuximab

(Erbitux) for colon cancer¹¹, and, on the other hand, EGFR-specific tyrosine kinase inhibitors

(TKIs) targeting the receptor catalytic domain, such as gefitinib (1, Iressa)¹²and erlotinib (2,

Tarceva)¹³for lung cancer (Figure 1)¹⁴, and dual inhibitors, such as lapatinib (EGFR/ErbB2)¹⁵

and vandetanib (EGFR/VEGFR, vascular endothelial growth factor receptor); others are

under clinical trials. However, somatic mutations in the kinase domain of EGFR induce drug

resistance^{14, 16, 17}. For example, erlotinib and gefitinib, which are effective in treating non–

small cell lung cancer tumors harboring a mutated form of EGFR, are ineffective against

18, 19

EGFR variants found in glioblastoma^8,

.

This, coupled with the fact that small

modifications to TKIs could have a strong influence in the binding mode and kinetics²⁰, has

encouraged sustained efforts to develop small-molecule ATP-competitive EGFR inhibitors

which target both wild type and mutated forms.

In this context, functionalized pyrido[2,3-d]pyrimidin-7(8H)-ones comprise a privileged

scaffold for pharmacologically active compounds with well-known activity as TKIs. More

particularly, 4-unsubstituted compounds of general structure 3 have shown IC₅₀in the range

M to nM in front of PDFGR, FGFR, EGFR, and c-Scr particularly when R¹and R⁴are aryl

groups (Figure 1)^21-23

.

3

Figure 1. Structures of gefitinib (1), erlotinib (2), 4-unsubstituted pyrido[2,3-d]pyrimidin-

7(8H)-ones (3), 4-amino-5,6-dihydropyrido[2,3-d]pyrimidin-7(8H)-ones (4), 4-

aminopyrido[2,3-d]pyrimidin-7(8H)-ones (5), and 2,6-dichlorophenyl substituted

pyridopyrimidines 6{1} and 7.

As part of our ongoing research in the area of TKIs we have described microwave assisted

synthetic methodologies to access 2-arylamino substituted 4-amino-5,6-dihydropyrido[2,3-

d]pyrimidin-7(8H)-ones (4, R¹and R⁴= aryl) (Ref.²⁴and references therein) and the

corresponding dehydrogenated compounds 5 (Figure 1)²⁵. Using such methodologies we

have recently described the synthesis of compound 7 active against non-Hodgkin´s

lymphomas (NHLs) by inhibiting the most upstream tyrosine kinases in the B cell receptor

(BCR) signaling pathway which are involved in the mature B cell neoplasms (Figure 1)²⁶. The

precursor 6{1} (Figure 1) of compound 7 also presented inhibitory activities against relevant

TKs including EGFR. Interestingly, 4-amino substituted pyridopyrimidines 6{1} and 7

presented very low cellular toxicity on normal cells, contrary to 4-unsubstituted compounds

3, thus showing the importance of the 4-amino substituent. Of note, compounds 6{1} and 7

bear at position C6 the 2,6-dichlorophenyl substituent that has been widely used in the field

of TKIs due to its favored fitting into the hydrophobic back pocket adjacent to the ATP

binding site present in different TKs.

Today, computational methods play a critical role in lead discovery and optimization^27-31

.

In this work, we investigate the suitability of a molecular dynamics (MD)-based end-point

binding free energy protocol to estimate the relative affinities of a series of congeneric

compounds for EGFR as a tool to guide chemical synthesis toward the most (in silico)

promising compounds. The MM/GBSA method used in this work has been successfully

evaluated to re-score docking poses³²(cf. also the applicability of the related MM/PBSA

method³³). While end-point methods based on classical mechanics are known to fail to

correctly estimate binding free energies in certain cases³⁴, MM/GBSA approaches were

successfully used to estimate relative binding free energies in several molecular systems^35-37

.

Starting with our EGFR model inhibitor 6{1}, its binding mode was assessed using

4

flexible molecular docking in dihedral coordinates and MD, in order to properly account for

protein flexibility³⁸; from there, a virtual combinatorial library around 6{1} was designed. In

order to investigate the validity of our methodology, selected analogues including some with

better and worse calculated binding affinities relative to 6{1} were synthesized, and their

biological activity determined. Analysis of the EGFR-analogue interaction in terms of

hydrogen bonding, van der Waals contribution, and water molecule bridging was undertaken

in order to understand the binding determinants of the different analogues. The experimental

validation was in good qualitative agreement with our theoretical calculations, while we also

obtained a 6-dibromophenyl-substituted molecule with improved performance toward EGFR

compared to the reference ligand.

2. Results and discussion

2.1. Characterization of the binding pose of ligand 6{1} and its interaction with EGFR.

The optimized structure of ligand 6{1} was docked into the ligand binding site of EGFR

using a two-stage protocol of rigid receptor docking followed by a full flexible docking

approach, similar to what has been done on other receptors^39-41(see Methods). The collected

poses were classified as “favorable” and “unfavorable”, according to their interaction pattern

(see Figure 2). It can be seen that the unfavorable pose has no hydrogen bonds with the

receptor, while the favorable pose of ligand 6{1} exhibits the following hydrogen bonds:

O(Gln791)∙∙∙HN(C4), Oγ1(Thr790)∙∙∙HN(C4), N3∙∙∙HN(Met793), and O(Met793)∙∙∙HN.

These residues are situated in the “hinge” region of the kinase, which connects the N and C

loves. Moreover, one of the chlorine atoms in the favorable pose makes a van der Waals

contact with the amide group of Ala743.

Figure 2. Favorable (magenta) and unfavorable (orange) poses of the ligand 6{1} within the

binding site (left panel). Interactions with the receptor for the favorable pose of ligand 6{1}

are shown on the right panel.

To further validate these results, and obtain a more detailed picture of the ligand-EGFR

interaction, molecular dynamics (MD) simulations were carried out on both poses, and the

relative binding energy estimated. An equilibration phase was performed, consisting of 1000

steps of minimization with the complex fixed, followed by 50 ps of NVT (0 to 300 K)

simulation for heating the system and water equilibration. Then, 150 ps of NPT (1 atm, 300

K) simulation was carried out to relax the box size. Finally, a production phase of 40 ns in a

5

NPT ensemble (1 atm, 300 K) was run. Figure 3 shows the RMSD for heavy atoms of 6{1} in

the “favorable” and “unfavorable” poses throughout the production phase (cf. Figure S1 for

the RMSD of the EGFR backbone atoms throughout the simulation). The calculated binding

energy using the MM/GBSA method of the “unfavorable” pose relative to the favorable one

was +20.5(0.1) kcal mol^-1, in full agreement with the qualitative analysis from the docking

results, and henceforth, only the “favorable” pose will be considered as starting point for the

simulations.

Figure 3. RMSD of “favorable” (blue) and “unfavorable” (red) poses of ligand 6{1}, in

complex with EGFR through 40 ns of the production phase (NPT ensemble, 1 atm, 300 K).

In order to compare the interaction of this ligand with other EGFR inhibitors, the EGFR-

6{1} complex was superimposed with some of the available experimental EGFR structures

complexed with other known ligands. Figure 4 shows the superposition of 6{1} to gefitinib

(PDB: 2ITY)⁴²and erlotinib (PDB: 1M17)⁴³bound to EGFR. The main fragments of these

ligands, as well as the aromatic rings with substituents, are placed in a similar location as in

ligand 6{1}, though it should be mentioned that those ligands they interact directly with the

“hinge” region by only one hydrogen bond with Met793.

Comparison of 6{1} and the

pyrazolo[3,4-d]pyrimidin-5-amine derivative described in Ref.⁴⁴also shows a similar binding

pattern.

Figure 4. Binding pose of ligand 6{1} (magenta color, receptor carbon atoms are in light

grey color) overlapped with erlotinib (PDB: 1M17, left) and gefitinib (PDB: 2ITY, right).

Residues Thr790, Gln791, and Met793 are shown in stick representation.

Figure 5 shows the hydrogen bond time evolution of ligand 6{1} with the hinge region

throughout the simulation. The hydrogen bonds N3∙∙∙HN(Met793), and O(Met793)∙∙∙HN

exhibit a conserved interaction throughout the simulation.

In the case of

O(Gln791)∙∙∙HN(C4), and Oγ1(Thr790)∙∙∙HN(C4) two main conformations are visited, since

the rotation around amino group, bounded to C4, is not restricted.

6

Figure 5. Time dependence of hydrogen bond (HB) distances between H atom and acceptor

atom of hydrogen bonds present in favorable pose of ligand 6{1} : N3∙∙∙HN(Met793)

(yellow), O(Met793)∙∙∙HN (red), O(Gln791)∙∙∙HN(C4) (blue), Oγ1(Thr790)∙∙∙HN(C4) (green).

In those bonds where the NH₂group is involved, distance was measured only for one of the

two H atoms.

2.2. In Silico design and characterization of ligand 6{1} analogues

With the aim to design ligands with enhanced activity toward EGFR, a chemical library

of small-molecule analogues was generated from ligand 6{x} (Table 1). These molecules

were generated by substitution of the chlorine atoms with synthetic feasible groups. Taking

into consideration that ring A is constrained to rotate within the binding site (Figure 6), in

those molecules where substituents R¹and R²were different, both combinations were

generated (labeled 6{Na} and 6{Nb}, respectively), assuming that R¹refers to the substituent

pointing toward the binding site.

Figure 6. Ligand 6{1} within the binding side of EGFR showing the hindered rotation of

aromatic ring A.

The complex between EGFR-6{1} obtained from the last frame of MD trajectory was

considered as the starting point to build protein-ligand complexes with molecules 6{2-11b}.

Then, 1000 steps of minimization were applied, followed by 20 ns of NPT (1 atm, 300K)

simulation. This protocol was also applied to ligand 6{1}. To estimate binding free energies,

we expected the MM/GBSA to be suitable for this case^{45, 46}, though other methods could be

also valid⁴⁷. On Table 1 the binding energies calculated using the MM/GBSA method are

listed (ΔG’), where given the structural similarity throughout the series, the ligand and

receptor entropic contributions upon binding were assumed constant, and thus have not been

included [cf. Eq. (1), Experimental Section]; it should be stressed that the entropic

7

contribution of the solvent is de facto accounted for in the solvation term. Following an

analysis of the MM/GBSA energy (electrostatic energy, electrostatic and nonpolar

contribution to the solvation energy, and van der Waals contribution), it could be seen that the

van der Waals energy term (ΔE_vdW, see Table 1) had the highest correlation with the binding

energy, a hint that the differential interaction with the receptor could be governed by that

contribution.

Based on these results, two compounds with better and worse predicted affinity than 6{1}

were selected to advance to synthesis: 6{2}, with R¹= R²= Br, and 6{10}, with R¹= F and

R²= CF₃. As a reference, 6{4}, with R¹= R²= H was also included, to shed light on the

impact of substituents on ring A.

Table 1. Binding energies (ΔG’, kcal mol^-1) and van der Waals contribution (ΔE_vdW, kcal

mol^-1) calculated using MM/GBSA. The values in parentheses represent the standard error of

the mean.

Compound R¹

R²

MM-GBSA

ΔE_vdW

ΔG’

6{1}

6{2}

6{3}

Cl

Br

F

H

Br

H

F

H

F

Cl

H

CF₃

F

Br

F

CF₃

F

OCH₃

Cl

Br

F

H

Br

H

F

Cl

F

CF₃

H

Br

F

CF₃

F

OCH₃

F

-51.2 (0.1)

-54.3 (0.1)

-48.1 (0.1)

-46.6 (0.1)

-49.6 (0.1)

-48.9 (0.1)

-46.5 (0.1)

-47.0 (0.1)

-47.6 (0.1)

-48.9 (0.1)

-49.5 (0.1)

-49.9 (0.1)

-50.7 (0.1)

-48.9 (0.1)

-48.6 (0.1)

-50.6 (0.1)

-50.1 (0.1)

-49.5 (0.1)

-54.6 (0.1)

-47.9 (0.1)

-44.9 (0.1)

-49.0 (0.1)

-47.3 (0.2)

-41.6 (0.1)

-47.2 (0.1)

-46.3 (0.1)

-49.5 (0.1)

-46.4 (0.1)

-45.3 (0.1)

-49.4 (0.1)

-48.0 (0.1)

-45.8 (0.1)

-47.3 (0.2)

-44.8 (0.2)

6{4}

6{5a}

6{5b}

6{6a}

6{6b}

6{7a}

6{7b}

6{8a}

6{8b}

6{9a}

6{9b}

6{10a}

6{10b}

6{11a}

6{11b}

2.3. Chemistry

8

As it is depicted in Scheme 1, the synthesis of compounds 6{x} consists of a 6 steps route

and the diversity of R¹and R²was introduced at the beginning through the synthesis of the

corresponding aryl acetate 8{x}. All the synthesis of these starting materials and their

corresponding 2-aryl acrylates 9{x} by condensation of 8{x} with paraformaldehyde in the

presence of K₂CO₃in DMF were described in a previous work⁴⁸. The 2-aryl acrylates 9{x}

were refluxed in MeOH with malononitrile (10) in the presence of NaOMe to afford through

a Michael addition the corresponding pyridones 11{x} in very good yields (up to 93%). Then,

pyridones 11{x} were condensed with phenyl guanidine (12) leading the corresponding

intermediates 13{x} which were transformed to the corresponding pyrido[2,3-d]pyrimidines

14{x} through a Dimroth rearrangement upon treatment with NaOMe/MeOH under

microwave irradiation at 140 ºC during 40 min. To accomplish the dehydrogenation of

compounds 14{x}, different procedures were carried out depending on the substituents

presents in R¹and R². When R¹= R²= Br, 14{2} was heated at 100 ºC for 4h in anhydrous

DMSO in presence of NaH to afford 15{2}, whereas when R¹= F and R²= CF₃, 14{10} was

refluxed with activated MnO₂in AcOH for 3h to yield 15{10}. Finally, the desired

compounds 6{x} were obtained by methylation of the lactam nitrogen of the corresponding

15{x} with MeI in DMSO in the presence of NaH, the yields being in the range 93–96%.

Scheme 1: Synthesis of pyrido[2,3-d]pyrimidines 6{x}

2.4. Biological activity assessment

After their synthesis, compounds 6{1}, 6{2}, 6{4} and 6{10} were evaluated at

Proqinase (http://www.proqinase.com) by measuring residual activity values (%) at a

concentration of 10 M in front of wild type EFGR. The results obtained (Table 2) indicate

the best inhibition activity is for compound 6{2} (R¹= R²= Br) followed by compound 6{1}

(R¹= R²= Cl) and finally the worst inhibitory activity is for 6{10} (R¹= F and R²= CF₃),

showing the following values of residual activity (% of control activity) 3%, 14% and 61%,

respectively. These experimental inhibitory data are in good qualitative agreement with our

calculations, showing that MM/GBSA calculations are useful for the purpose used in this

work. These results would seem indicate that the bulkier the substituent R¹and R², the more

active is the compound 6{x}. This observation is supported by the almost inactivity observed

for compound 6{4} (R¹= R²= H) but the required bulkiness has a limit as is clearly shown

by compound 6{10} (R¹= F and R²= CF₃).

9

Table 2. Residual activity values (%) at a concentration of 10 M of pyrido[2,3-

d]pyrimidines 6{1}, 6{2}, 6{4} and 6{10} in front of EGFR wild type.

Structure

6{1}

Residual activity (%)

14

3

6{2}

6{4}

6{10}

49

61

2.5. Analysis of EGFR-6{x} interaction

Analysis of the RMSD for each of the ligands shows that the gly-rich loop, the activation

loop, the αC helix and the DFG motif are stable throughout the simulation. Regarding

oscillations around the average conformation, it was only observed that the gly-rich loop

exhibits a larger RMSF in the case of ligand 6{10} (see RMSF analysis in Supplementary

Information).

In order to investigate the influence of the hydrogen bonding network on binding

energies, the fraction of hydrogen bonding during the MD simulations was calculated with

cpptraj included in AmberTools13. The results are displayed in Table 3 , where no substantial

difference can be observed in those patterns.

10

Table 3. Donor and acceptor hydrogen bond atoms between ligands and EGFR residues

through the last 10 ns of simulation. For distance cut-off and angular cut-off, values of 3.4Å

and 150º were selected, respectively. Average distances between heavy atoms (Å), and angles

(º) are also shown. The values in parentheses represent the standard deviation. The two

hydrogen atoms of NH₂group are discriminated with apostrophe on one of them.

6{1}

6{2}

6{4}

Acceptor

N3

Oγ1(Thr790)

O(Met793)

Oγ1(Thr790)

O(Gln791)

Donor

HN(Met793)

HN(C4)

HN

H’N(C4)

Fraction <Distance> <Angle>

0.86

0.30

0.23

0.02

3.10 (0.16) 164.9 (9.6)

3.07 (0.15) 162.5 (7.0)

2.93 (0.14) 156.8 (5.3)

3.08 (0.15) 162.7 (7.0)

2.89 (0.14) 154.1 (3.7)

Acceptor

N3

Oγ1(Thr790)

O(Met793)

O(Gln791)

Oγ1(Thr790)

Donor

HN(Met793)

HN(C4)

HN

H’N(C4)

Fraction <Distance> <Angle>

0.79

0.72

0.33

0.08

0.05

3.09 (0.13) 164.1 (7.0)

2.97 (0.13) 163.0 (7.1)

2.92 (0.14) 156.6 (5.1)

2.91 (0.14) 154.5 (4.0)

2.96 (0.13) 163.3 (7.7)

Acceptor

N3

Donor

HN(Met793)

HN

HN(C4)

H’N(C4)

HN(C4)

H’N(H)

Fraction <Distance> <Angle>

0.83

0.37

0.23

0.19

0.04

0.03

0.02

3.09 (0.14) 164.1 (7.0)

2.91 (0.13) 157.7 (5.7)

3.07 (0.16) 162.7 (7.0)

3.06 (0.15) 163.3 (7.2)

3.24 (0.11) 155.7 (5.0)

2.84 (0.12) 155.3 (4.3)

2.89 (0.14) 154.0 (4.1)

O(Met793)

Oγ1(Thr790)

N(Met793)

O(Gln791)

6{10a}

Acceptor

N3

Donor

HN(Met793)

HN(C4)

HN

H’N(C4)

HN(C4)

Fraction <Distance> <Angle>

0.81

0.46

0.41

0.24

0.05

0.03

3.12 (0.13) 164.1 (6.9)

3.09 (0.15) 163.9 (7.0)

2.97 (0.15) 157.2 (5.3)

3.08 (0.15) 163.7 (7.2)

2.84 (0.12) 154.9 (4.1)

2.84 (0.12) 154.3 (4.0)

Oγ1(Thr790)

O(Met793)

Oγ1(Thr790)

O(Gln791)

Decomposition of the binding energy and the van der Waals interaction between

compounds and their neighboring residues are plotted in Figure 7. Overall, both plots follow

a similar pattern in almost the entire range, which is consistent with the hypothesis that the

van der Waals term has a predominant effect on the binding energy. Residues with the highest

stabilizing contributions are Ala743, Lys745, Thr790, Met793, and Leu844, where the first

three are close to the substituted aromatic ring A. The differences between the distances

R¹∙∙∙O(Ala743) and the van der Waals radii of the atoms involved are reported on Table 4.

This magnitude increases in the order 6{2} < 6{1} < 6{10a} < 6{4}, what correlates with the

loss of binding affinity. A similar effect is also observed between R¹and the N atom of the

adjacent residue Ile744. The linear correlation coefficients between components and binding

energies shown in Table S1 confirm this trend.

11

Figure 7. Decomposition of the binding energies ΔG’ (left) and the van der Waals

contribution ΔE_vdW(right) on a pairwise energy decomposition scheme for 6{1} (blue), 6{2}

(green), 6{4} (yellow), and 6{10a} (red).

Table 4. Mean bond distance σ between R¹substituent and the carbonyl oxygen of Ala743,

standard deviation in parenthesis, and the difference δ between σ and the van der Waals radii

of atoms involved [δ = σ - r_vdw(O) - r_vdw(R¹)] (Distances in Å).

R¹…O(Ala743)

σ

δ

6{2} Br∙∙∙O 3.48 (0.21)

6{1} Cl∙∙∙O 3.46 (0.26)

6{10a} F∙∙∙O 3.28 (0.30)

6{4} H∙∙∙O 3.65 (0.41)

0.11

0.19

0.29

0.93

Analysis of water clusters on all four ligands shows a conserved O=C(7) interaction

through hydrogen bonding with a water molecule. Water distribution around O=C(7) is

similar for all ligands with the exception of the 6{10a} complex, where the displacement of

Lys745 restricts the available space. There are other two water molecules within the binding

site interacting with Gln791 and Thr854, however, unlike the case of erlotinib^{43, 49}, these

water molecules are not directly involved in hydrogen bonding between the ligand and

EGFR. In the case of 6{1} there is another water molecule at the bottom of the pocket,

which interacts with Phe856.

3. Conclusions

Currently, most hit-to-lead campaigns are based on a “trial and error” approach. A

computationally-aided synthetic approach is usually not undertaken. In this work we

explored the use of a molecular dynamics-based end-point binding free energy protocol to

estimate the relative affinities of a virtual combinatorial library designed around a 4-amino

substituted pyridopyrimidine EGFR model inhibitor (6{1}) as a tool to guide chemical

synthesis toward the most promising compounds, and then validate our predictions through

biological activity evaluation.

It was observed that the experimental inhibition activity was in good qualitative

agreement with our binding free energy calculations using the MM/GBSA method for the

four synthesized compounds. From the MD trajectories, we concluded that differences at the

binding free energy level are governed by van der Waals interaction, showing the compounds

examined negligible differences in terms of hydrogen bonding, and water molecules bridging

their interaction with EGFR. It should be also mentioned that a 6-dibromophenyl-substituted

molecule which exhibited an enhanced affinity toward EGFR compared to the reference

12

ligand was obtained. Although further validation is needed on other molecular systems –

since in many cases the usefulness of a given computational method is system-dependent-, we

consider that this approach offers an attractive balance of performance and computational

affordability.

4. Experimental section

4.1. Docking

Ligand 6{1} was docked within the binding site of EGFR (PDB: 2ITW⁴²) using a

flexible-ligand/rigid-receptor approach, as implemented in the ICM package^{50, 51}. Two

dominant ligand binding modes were observed, and a representative complex of each was

chosen, and subjected to a flexible-ligand/flexible-side chain Monte Carlo-based global

energy optimization^{40, 52-55}. Solvation effects were taken into account using a Generalized Born

model⁵⁶. A conformational stack was collected throughout the simulation of each complex⁵⁷,

and the best-energy structure corresponding to each initial binding mode was kept, labeled as

“favorable” and “unfavorable”, according to their relative energy.

4.2. Preparation of the molecular systems

The simulations were based on the X-ray crystal structure of EGFR in complex with

(PDB: 2ITW)⁴². In the preparation of the receptor, all Asp and Glu residues were considered

to have a negative charge and all the Arg and Lys residues were considered to have a positive

charge. Histidine tautomers were assigned following the hydrogen bonding pattern. In the

case of His835, close to binding site, was protonated on the Nε.

4.3. Molecular dynamics (MD) simulations

The complexes have a net charge of +1. To achieve electroneutrality, a chloride was

added as counterion, with the Leap module. The neutralized complexes were immersed in a

box of TIP3P⁵⁸waters which extended up to 10 Å from the solute. The protein was described

using the Amber99SB force field⁵⁹with the dielectric constant taken as 1, while the ligands

were described using the Generalized Amber Force Field (GAFF)⁶⁰, with charges derived

from AM1-BCC⁶¹, which were calculated with the Antechamber module. Leap and

Antechamber are included in the package AmberTools 13.0⁶².

All MD simulations were run using the NAMD 2.9 software⁶³. The van der Waals

interaction cutoff distances were set at 12 Å and long-range electrostatic forces were

computed using the particle mesh Ewald summation method with a grid size set to 1.0 Å. The

1-4 contributions were multiplied by a factor of 0.83 to match the AMBER force field

requirements. For all production simulations, constant temperature (300 K) was maintained

using Langevin dynamics with a damping coefficient of 5 ps^-1, while pressure was kept

constant at 1 atm through the Nosé-Hoover Langevin piston method with a decay period of

200 fs and a damping time constant of 100 fs. A time step of 1 fs was used along molecular

mechanics. Bonds involving hydrogen atoms were constrained using the SHAKE algorithm⁶⁴.

4.4. MM/GBSA calculations

Small-molecule-protein binding free energies were computed using the MM/GBSA

method for all complexes, where the binding free energy is calculated as the difference

between the bound and unbound states of protein and ligand, according to^{34, 65-68}

13

G_bind  E_MM G_solvTS_bind

 G^'TS_bind

(1)

where … is the trajectory average, E_MMis the gas-phase potential energy, and ΔG_solv= ΔW

(W is the effective solvation energy which incorporates the solvent degrees of freedom).

Provided the high similarity of the molecules on Table 1, the entropic changes upon binding

were assumed constant, and thus ΔG’ is reported on Table 1. The solvation free energy ΔG_solv

was separated into polar and non-polar contributions as

pol

solv

np

solv

(2)

G_solv G  G

The polar contribution to the solvation free energy was calculated using the generalized Born

(GB) model⁶⁷implemented in MMPBSA.py module of Amber⁶⁹, igb=2 as selected model.

The hydrophobic contribution to the solvation free energy ( G_nonp) was determined using the

solvent-accessible surface area (SASA) as

G_nonp  SASA  

(3)

where values for γ and β were set to 0.0072 kcal·mol⁻²and 0 kcal·mol⁻¹, respectively. The

protein–ligand binding free energy was calculated using a single trajectory (for ligand,

receptor and complex)^{34, 45}based on 1000 snapshots taken from the last 10 ns portion (10 ps

interval) of the MD simulation trajectories.

For the purpose of obtaining the detailed representation of the ligands/EGFR interactions,

free energy decomposition analysis was employed to decompose the total binding free

energies into ligand–residue pairs. These calculations were performed using a pairwise

energy decomposition scheme (idecomp option 3) also with the MMPBSA.py module. In this

scheme, interactions are decomposed by specific residue pairs by including only those

interactions in which one atom from each of the analyzed residues is participating, following

the work of Gohlke et al.⁷⁰.

4.5. Synthesis and characterization of compounds

4.5.1. General

All solvents and chemicals were reagent grade. Unless otherwise mentioned, all solvents

and chemicals were purchased from commercial vendors (Fluka, Aldrich, ABCR and ACROS

Organics) and used without purification. Compound 6{1} was prepared as previously

1

13

described²⁶. H and C NMR spectra were recorded on a Varian 400-MR spectrometer that

was operating at a field strength of 400 and 100.6 MHz, respectively. Chemical shifts were

reported in parts per million () and coupling constants (J) were in Hz by using, in the case of

¹H NMR spectroscopy, TMS as an internal standard, and in the case of ¹³C NMR

spectroscopy the solvent at 39.5 ppm (DMSO-d₆) or at 29.84 ppm (acetone-d₆) as an internal

reference. Standard and peak multiplicities are designed as follows: s, singlet; d, doublet; t,

triplet; q, quartet; br, broad signal. IR spectra were recorded in a Thermo Scientific Nicolet

iS10 FTIR spectrophotometer with Smart iTr. Wavenumbers () are expressed in cm^-1. MS

data (m/z (%), EI, 70 eV) were obtained by using an Agilent Technologies 5975 spectrometer

and a Hewlett Packard HP5988A quadrupole mass spectrometer operating in electronic

ionization (EI) mode at 70 eV and at 4 kV accelerating potential, or a Bruker Biotoff II

spectrometer operating in electrospray ionization (ESI) mode with a Time of Flight (TOF)

detector or on a VG AutoSpec (Micromass Instruments). HRMS data were obtained by using

a VG AutoSpec (Micromass Instruments) Trisector EBE high resolution spectrometer (EI

mode), a Bruker Biotof II mass spectrometer (ESI TOF mode). Elemental microanalyses

14

were obtained on a EuroVector Instruments Euro EA elemental analyzer. The melting points

were determined with a Büchi-Tottoli 530 capillary apparatus and are uncorrected.

Microwave irradiation experiments were carried out in a Initiator^TM(Biotage) microwave

apparatus, operating at a frequency of 2.45 GHz with continuous irradiation power from 0 to

400 W. Reactions were carried out in 0.5, 2.5, 5, 20 mL glass tubes, sealed with

o

aluminium/Teflon crimp tops, which can be exposed up to 250 C and 20 bar internal

pressure. Temperature was measured with an IR sensor on the outer surface of the process

o

vial. After the irradiation period, the reaction vessel was cooled rapidly to 50 C by air jet

cooling.

The synthesized compounds have been checked for their melting points, physical nature,

IR, ¹H NMR, ¹³C NMR, Mass spectroscopy and Elemental analysis for individual compounds

and the data are summarized as under

4.5.2. 5-(2,6-dibromophenyl)-2-methoxy-6-oxo-1,4,5,6-tetrahydropyridine-3-carbonitrile

(11{2})

To a solution of 913.0 mg (16.42 mmol) of NaOMe in 20 mL of anhydrous methanol,

802.0 mg (12.14 mmol) of malononitrile (10) were added and the mixture left cool down.

3.26 g (10.2 mmol) of methyl 2-(2,6-dibromophenyl)acrylate (9{2})⁴⁸were added slowly and

the mixture refluxed for 5 h. The solvent was evaporated in vacuo and the residue dissolved

in the minimum quantity of water. Careful neutralization to pH 7 with 2M aqueous HCl

allowed the precipitation of a solid which was filtered, washed with cold water and dried in

vacuo over phosphorus pentoxide. 3.750 g (9.71 mmol, 95%) of 5-(2,6-dibromophenyl)-2-

methoxy-6-oxo-1,4,5,6-tetrahydropyridine-3-carbonitrile (11{2}) were obtained as a white

solid, mp 200-203 ^oC. ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 10.88 (br s, 1H), 7.71 (ddd, J

= 16.1, 8.0, 1.0 Hz, 2H), 7.19 (t, J = 8.0 Hz, 1H), 4.71 (dd, J = 13.9, 8.8 Hz, 1H), 3.97 (s,

3H), 2.97 (dd, J = 15.2, 13.9 Hz, 1H), 2.46 (dd, J = 15.2, 8.8 Hz, 1H). ¹³C NMR (100.6 MHz,

DMSO-d₆) δ (ppm): 169.0, 159.5, 136.5, 133.9, 132.3, 130.8, 126.5, 124.2, 118.3, 62.5, 59.1,

47.3, 25.1. IR (KBr): (cm^-1): 3230, 3183, 2926, 2203, 1713, 1645, 1487, 1254. MS (70 eV,

EI): m/z (%) = 385.8 (19) [M]⁺, 304.9 (12) [M-Br]⁺, 275.7 (100) [M-CH₃OBr]⁺. HRMS (70

eV, EI): calcd. for C₁₃H₁₀Br₂N₂O₂: 383.9109 [M]⁺, found: 383.9109.

4.5.3. 2-methoxy-6-oxo-5-phenyl-1,4,5,6-tetrahydropyridine-3-carbonitrile (11{4})

As above for 11{2} but using NaOMe (1.289 g, 23.18 mmol), anhydrous methanol (29

mL), malononitrile (10) (1.133 g, 17.15 mmol) and methyl 2-phenylacrylate (9{4})⁴⁸(2.34 g,

14.4 mmol) to afford 1.741 g (7.63 mmol, 53%) of 2-methoxy-6-oxo-5-phenyl-1,4,5,6-

o

1

tetrahydropyridine-3-carbonitrile (11{4}) as a yellow solid, mp 138-141 C. H NMR (400

MHz, DMSO-d₆) δ (ppm): 10.78 (br s, 1H), 7.37 – 7.32 (m, 2H), 7.31 – 7.27 (m, 1H), 7.26 –

7.23 (m, 2H), 3.95 (s, 3H), 3.85 (dd, J = 11.0, 6.7 Hz, 1H), 2.73 (dd, J = 15.4, 11.0 Hz, 1H),

13

2.60 (dd, J = 15.4, 6.7 Hz, 1H). C-RMN (100.6 MHz, DMSO-d₆) δ (ppm): 171.3, 159.7,

137.8, 128.37, 128.36, 127.1, 118.5, 63.2, 58.9, 45.5, 28.0. IR (KBr): (cm^-1): 3427, 3226,

2204, 1712, 1644. MS (ESI-TOF): m/z (%) = 229.1 (20) [M+H]⁺. HRMS (ESI-TOF): calcd.

for C₁₃H₁₃N₂O₂: 229.0972 [M+H]⁺, found: 229.0969.

4.5.4.

5-(2-fluoro-6-(trifluoromethyl)phenyl)-2-methoxy-6-oxo-1,4,5,6-

tetrahydropyridine-3-carbonitrile (11{10})

As above for 11{2} but using NaOMe (913.0 mg, 16.42 mmol), anhydrous methanol (20

mL), malononitrile (10) (802.0 mg, 12.14 mmol) and methyl 2-(2-fluoro-6-

(trifluoromethyl)phenyl)acrylate (9{10}) (2.53 g, 10.2 mmol) to afford 2.969 g (9.45 mmol,

93%) of 5-(2-fluoro-6-(trifluoromethyl)phenyl)-2-methoxy-6-oxo-1,4,5,6-tetrahydropyridine-

o

1

3-carbonitrile (11{10}) as a white solid, mp 207-208 C. H NMR (400 MHz, DMSO-d₆) δ

15

(ppm): 10.93 (br s, 1H), 7.67 – 7.57 (m, 3H), 4.14 (dd, J = 13.8, 7.1 Hz, 1H), 3.98 (s, 3H),

13

2.79 (t, J = 15.0 Hz, 1H), 2.43 (dd, J = 15.3, 7.5 Hz, 1H). C NMR (100.6 MHz, DMSO-d₆)

δ (ppm): 169.2, 161.0 (d, J = 248.1 Hz), 159.85, 130.4 (d, J = 9.7 Hz), 129.7 (dq, J = 30.2,

5.6 Hz), 124.3 (d, J = 16.0 Hz), 123.6 (dq, J = 274.3, 3.6 Hz), 121.9, 120.7 (d, J = 22.6 Hz),

118.2, 62.9, 59.0, 40.15, 26.6 (d, J = 3.3 Hz). IR (KBr): (cm^-1): 3220, 3180, 2960, 2199,

1715, 1639, 1487, 1323, 1251, 1120. MS (70 eV, EI): m/z (%) = 314.0 (16) [M]⁺, 204.0 (100)

[M-C₃H₃F₃N]⁺. HRMS (70 eV, EI): calcd. for C₁₄H₁₀F₄N₂O₂: 314.0678 [M]⁺, found:

314.0680.

4.5.5.

2-amino-6-(2,6-dibromophenyl)-4-imino-3-phenyl-4,5,6,8-tetrahydropyrido[2,3-

d]pyrimidin-7(3H)-one (13{2})

A mixture of N-phenylguanidine carbonate (12 having a C₇H₉N₃・(H₂CO₃)_0.7

stoichiometry) (523.0 mg, 3.07 mmol of N-phenylguanidine), sodium methoxide (232.5 mg,

4.30 mmol), and 1,4-dioxane (15 mL) is sealed in a 20 mL microwave vial and heated at 65

ºC under microwave irradiation for 15 min. A clear solution with a white precipitate is

obtained. The solid is removed by filtration and the mother liquor is transferred to a 20 mL

microwave vial together with 5-(2,6-dibromophenyl)-2-methoxy-6-oxo-1,4,5,6-tetra-

hydropyridine-3-carbonitrile (11{2}) (393.0 mg, 1.02 mmol). The vial is sealed and heated at

140 ºC under microwave irradiation for 40 min. The solvent of the red solution obtained is

removed in vacuo, and the resulting red oil is treated with acetone (10 mL) and sonication for

10 min while a white precipitate is formed. The solid is filtered, washed with cold acetone to

afford 402.4 mg (3.29 mmol) (81%) of 2-amino-6-(2,6-dibromophenyl)-4-imino-3-phenyl-

4,5,6,8-tetrahydropyrido[2,3-d]pyrimidin-7(3H)-one (13{2}) as a light orange solid, mp 262-

o

1

264 C. H NMR (400 MHz, DMSO-d₆) δ (ppm): 10.00 (br s, 1H), 7.72 (dd, J = 8.0, 1.1 Hz,

1H), 7.69 (dd, J = 8.0, 1.1 Hz, 1H), 7.58 (t, J = 7.5 Hz, 2H), 7.55 – 7.48 (m, 1H), 7.32 (d, J =

7.3 Hz, 2H), 7.18 (t, J = 8.0 Hz, 1H), 6.19 (br s, 2H), 5.54 (br s, 1H), 4.62 (dd, J = 13.3, 9.4

13

Hz, 1H), 2.86 (dd, J = 16.0, 9.4 Hz, 1H), 2.74 (dd, J = 15.9, 13.4 Hz, 1H). C NMR (100.6

MHz, DMSO-d₆) δ (ppm): 169.7, 156.3, 153.9, 149.0, 138.2, 135.3, 135.2, 133.8, 132.1,

130.5, 130.4, 129.5, 129.4, 129.1, 126.3, 124.3, 84.9, 47.9, 24.5. IR (KBr): (cm^-1): 3396,

3174, 1685, 1634, 1525, 1486, 1430, 1379, 1315, 1268, 1198, 767, 703. Anal. calcd for

C₁₉H₁₅N₅OBr₂: C: 46.65%, H: 3.09%, N: 14.32%, O: 3.27%, Br: 32.67%; Found: C: 46.49%,

H: 2.99%, N: 14.23%. HRMS (ESI-TOF): calcd. for C₁₉H₁₆Br₂N₅O: 487.9716 [M+H]⁺,

found: 487.9718.

4.5.6.

one (13{4})

As above for 13{2} but using N-phenylguanidine carbonate (12 having a

2-amino-4-imino-3,6-diphenyl-4,5,6,8-tetrahydropyrido[2,3-d]pyrimidin-7(3H)-

C₇H₉N₃・(H₂CO₃)_0.7stoichiometry) (523.0 mg, 3.07 mmol of N-phenylguanidine), sodium

methoxide (232.5 mg, 4.30 mmol), dioxane (15 mL) and 2-methoxy-6-oxo-5-phenyl-1,4,5,6-

tetrahydropyridine-3-carbonitrile (11{4}) (232,4 mg, 1.02 mmol) to afford 192.7 mg (0.58

mmol, 53%) of 2-amino-4-imino-3,6-diphenyl-4,5,6,8-tetrahydropyrido[2,3-d]pyrimidin-

o

1

7(3H)-one (13{4}) as a white solid, mp 248-251 C. H NMR (400 MHz, DMSO-d₆) δ

(ppm): 9.94 (br s, 1H), 7.63 – 7.46 (m, 3H), 7.38 – 7.18 (m, 7H), 6.11 (br s, 2H), 5.04 (br s,

1H), 3.80 (dd, J = 9.4, 7.4 Hz, 1H), 2.84 (dd, J = 16.1, 7.3 Hz, 1H), 2.69 (dd, J = 15.9, 9.3

Hz, 1H). ¹³C NMR (100.6 MHz, DMSO-d₆) δ (ppm): 172.0, 153.7, 149.9, 139.7, 135.3,

130.5, 129.5, 129.3, 129.1, 128.2, 128.2, 126.7, 86.2, 46.1, 26.4. IR (KBr): (cm^-1): 3491,

3412 3313, 3153, 2890, 1684, 1637, 1524, 1489, 1454, 1380, 1313, 1260, 1197, 770, 701.

HRMS (ESI-TOF): calcd. For C₁₉H₁₈N₅O: 332.1506 [M+H]⁺, found: 332.1514.

16

4.5.7.

2-amino-6-(2-fluoro-6-(trifluoromethyl)phenyl)-4-imino-3-phenyl-4,5,6,8-

tetrahydropyrido[2,3-d]pyrimidin-7(3H)-one (13{10})

As above for 13{2} but using N-phenylguanidine carbonate (12 having a

C₇H₉N₃・(H₂CO₃)_0.7stoichiometry) (523.0 mg, 3.07 mmol of N-phenylguanidine), sodium

methoxide (232.5 mg, 4.30 mmol), dioxane (15 mL) and 5-(2-fluoro-6-

(trifluoromethyl)phenyl)-2-methoxy-6-oxo-1,4,5,6-tetrahydropyridine-3-carbonitrile

(11{10}) (320.0 mg, 1.02 mmol) to afford 393.9 mg (0.94 mmol, 93%) of 2-amino-6-(2-

fluoro-6-(trifluoromethyl)phenyl)-4-imino-3-phenyl-4,5,6,8-tetrahydropyrido[2,3-

o

1

d]pyrimidin-7(3H)-one (13{10}) as a white solid, mp 252-254 C. H NMR (400 MHz,

DMSO-d₆) δ (ppm): 10.05 (br s, 1H), 7.70 – 7.45 (m, 6H), 7.31 (dd, J = 13.2, 7.5 Hz, 2H),

6.16 (br s, 2H), 5.27 (br s, 1H), 4.01 (dd, J = 13.8, 8.1 Hz, 1H), 2.92 (dd, J = 15.3, 8.2 Hz,

13

1H), 2.57 (dd, J = 15.0, 13.8 Hz, 1H). C NMR (100.6 MHz, DMSO-d₆) δ (ppm): 169.7,

161.1 (d, J = 247.8 Hz), 156.0, 153.9, 149.7, 135.2, 130.6, 129.9 (d, J = 9.8 Hz), 129.7 (m),

129.5, 129.4 (m), 129.4, 129.2, 126.5 (d, J = 16.5 Hz), 123.7 (dq, J = 274.1, 3.2 Hz), 121.7

(m), 120.6 (d, J = 22.9 Hz), 85.7, 40.7, 25.8 (d, J = 2.6 Hz). IR (KBr): (cm^-1): 3446, 3312,

3177, 2944, 1689, 1643, 1523, 1468, 1384, 1318, 1267, 1167, 1120, 807. HRMS (ESI-TOF):

calcd. for C₂₀H₁₆F₄N₅O: 418.1285 [M+H]⁺, found: 418.1300.

4.5.8.

d]pyrimidin-7(6H)-one (14{2})

mixture of

4-amino-6-(2,6-dibromophenyl)-2-(phenylamino)-5,8-dihydropyrido[2,3-

A

2-amino-6-(2,6-dibromophenyl)-4-imino-3-phenyl-4,5,6,8-

tetrahydropyrido[2,3-d]pyrimidin-7(3H)-one (13{2}) (1.609 g, 3.29 mmol), sodium

methoxide (177.7 mg, 3.29 mmol) and methanol (10 mL) is sealed in a 20 mL microwave

vial and heated at 140 ºC under microwave irradiation for 40 min. The white solid obtained is

filtered, washed with water, ethanol and diethyl ether to afford 1.429 mg (2.92 mmol) (89%)

of

pure

4-amino-6-(2,6-dibromophenyl)-2-(phenylamino)-5,8-dihydropyrido[2,3-

1

o

d]pyrimidin-7(6H)-one (14{2}), mp > 280 C. H NMR (400 MHz, DMSO-d₆) δ (ppm):

10.35 (br s, 1H), 8.75 (br s, 1H), 7.84 (dd, J = 8.7, 1.0 Hz, 2H), 7.73 (ddd, J = 12.6, 8.0, 1.2

Hz, 2H), 7.23 – 7.15 (m, 3H), 6.84 (tt, J = 7.3, 1.1 Hz, 1H), 6.39 (br s, 2H), 4.70 (dd, J =

13.0, 9.2 Hz, 1H), 2.96 (dd, J = 15.9, 9.2 Hz, 1H), 2.85 (dd, J = 15.9, 13.0 Hz, 1H). ¹³C NMR

(100.6 MHz, DMSO-d₆) δ (ppm): 169.2, 161.5, 158.3, 155.6, 141.4, 138.2, 133.9, 132.1,

130.5, 128.2, 126.3, 124.4, 120.2, 118.4, 84.1, 47.9, 23.7. IR (KBr): (cm^-1): 3499, 3400,

3286, 3201, 3136, 3090, 2917, 1676, 1637, 1612, 1574, 1550, 1499, 1478, 1441, 1382, 1246,

778, 757. Anal. calcd for C₁₉H₁₅N₅OBr₂: C: 46.65%, H: 3.09%, N: 14.32%, O: 3.27%, Br:

32.67%; Found: C: 46.67%, H: 2.90%, N: 14.18%. HRMS (ESI-TOF): calcd. for

C₁₉H₁₆Br₂N₅O: 487.9716 [M+H]⁺, found: 487.9713.

4.5.9. 4-amino-6-phenyl-2-(phenylamino)-5,8-dihydropyrido[2,3-d]pyrimidin-7(6H)-one

(14{4})

As

above

for

14{2}

but

using

2-amino-4-imino-3,6-diphenyl-4,5,6,8-

tetrahydropyrido[2,3-d]pyrimidin-7(3H)-one (13{4}) (497.1 mg, 1.5 mmol), sodium

methoxide (81.0 mg, 1.5 mmol) and methanol (5 mL) to afford 416.0 mg (1.26 mmol, 84%)

of pure 4-amino-6-phenyl-2-(phenylamino)-5,8-dihydropyrido[2,3-d]pyrimidin-7(6H)-one

o

(14{4}) as a white solid, mp > 280 C. ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 10.27 (br s,

1H), 8.72 (br s, 1H), 7.83 (d, J = 7.7 Hz 2H), 7.36 – 7.29 (m, 2H), 7.29 – 7.22 (m, 3H), 7.18

(t, J = 7.5 Hz, 2H), 6.84 (t, J = 7.3 Hz, 1H), 6.38 (br s, 2H), 3.87 (dd, J = 9.3, 7.0 Hz, 1H),

2.95 (dd, J = 15.9, 6.9 Hz, 1H), 2.82 (dd, J = 15.9, 9.5 Hz, 1H). ¹³C NMR (100.6 MHz,

DMSO-d₆) δ (ppm): 172.0, 161.3, 158.2, 156.1, 141.5, 139.6, 128.3, 128.2, 128.2, 126.8,

120.2, 118.4, 85.7, 46.2, 25.6. IR (KBr): (cm^-1): 3469, 3255, 3171, 2926, 1696, 1642, 1591,

17

1574, 1542, 1485, 1449, 1433, 1379, 1243, 783, 785, 700. HRMS (ESI-TOF): calcd.

forC₁₉H₁₈N₅O: 332.1506 [M+H]⁺, found: 332.1508.

4.5.10.

4-amino-6-(2-fluoro-6-(trifluoromethyl)phenyl)-2-(phenylamino)-5,8-

dihydropyrido[2,3-d]pyrimidin-7(6H)-one (14{10})

As above for 14{2} but using 2-amino-6-(2-fluoro-6-(trifluoromethyl)phenyl)-4-imino-3-

phenyl-4,5,6,8-tetrahydropyrido[2,3-d]pyrimidin-7(3H)-one (13{10}) (1.576 g, 3.78 mmol),

sodium methoxide (203.9 mg, 3.78 mmol) and methanol (10 mL) to afford 1.261 g (3.02

mmol, 80%) of pure 4-amino-6-(2-fluoro-6-(trifluoromethyl)phenyl)-2-(phenylamino)-5,8-

1

o

dihydropyrido[2,3-d]pyrimidin-7(6H)-one (14{10}) as a white solid, mp > 280 C. H NMR

(400 MHz, DMSO-d₆) δ (ppm): 10.45 (br s, 1H), 8.78 (br s, 1H), 7.90 – 7.78 (m, 2H), 7.72 –

7.58 (m, 3H), 7.26 – 7.14 (m, 2H), 6.85 (tt, J = 7.5, 1.1 Hz, 1H), 6.42 (br s, 2H), 4.09 (dd, J =

13.6, 7.7 Hz, 1H), 3.02 (dd, J = 16.3, 7.7 Hz, 1H), 2.68 (t, J = 14.8 Hz, 1H). ¹³C NMR (100.6

MHz, DMSO-d₆) δ (ppm): 169.5, 161.21, 161.18 (d, J = 248.0 Hz), 158.4, 155.9, 141.4,

130.0 (d, J = 9.7 Hz), 129.6 (dq, J = 29.5, 5.2 Hz), 128.22, 126.2 (d, J = 15.5 Hz), 123.7 (dq,

J = 274.2, 3.5 Hz), 121.7, 120.6 (d, J = 22.3 Hz), 120.3, 118.5, 84.9, 40.7, 24.9 (d, J = 2.5

Hz). IR (KBr): (cm^-1): 3472, 3249, 3204, 3119, 2934, 1692, 1640, 1589, 1573, 1545, 1497,

1468, 1434, 1386, 1320, 1256, 1112, 784. Anal. calcd for C₂₀H₁₅N₅OF₄: C: 57.56%, H:

3.62%, N: 16.78%, O: 3.83%, Br: 18.21%; Found: C: 57.43%, H: 3.84%, N: 16.64%. HRMS

(ESI-TOF): calcd. for C₂₀H₁₆F₄N₅O: 418.1285 [M+H]⁺, found: 418.1289.

4.5.11. 4-amino-6-(2,6-dibromophenyl)-2-(phenylamino)pyrido[2,3-d]pyrimidin-7(8H)-

one (15{2})

A mixture of 244.6 mg (0.5 mmol) of 4-amino-6-(2,6-dibromophenyl)-5,6-dihydro-2-

(phenylamino)pyrido[2,3-d]pyrimidin-7(8H)-one (14{2}) and 60.0 mg (1.5 mmol) of sodium

hydride (NaH) (60% dispersion in mineral oil) in 5 mL of anhydrous DMSO was heated for 4

hours at 100 ºC protected from moisture. The resulting solution was cooled down, water (300

mL) was added and it was neutralized with AcOH. The resulting precipitate was filtered,

washed with EtOH and EtOEt and dried in vacuo over phosphorus pentoxide to afford 199.5

mg (0.41 mmol, 82%) of 4-amino-6-(2,6-dibromophenyl)-2-(phenylamino)pyrido[2,3-

o

1

d]pyrimidin-7(8H)-one (15{2}) as a brownish solid, mp > 280 C. H NMR (400 MHz,

DMSO-d₆) δ (ppm): 11.85 (br s, 1H), 9.27 (br s, 1H), 8.00 (s, 1H), 7.90 (dd, J = 8.7, 1.0 Hz,

2H), 7.76 (d, J = 8.1 Hz, 2H), 7.31 (br s, 2H), 7.29 – 7.22 (m, 3H), 6.94 (t, J = 7.3 Hz, 1H).

¹³C NMR (100.6 MHz, DMSO-d₆) δ (ppm): 161.1, 161.0, 159.5, 156.1, 140.6, 138.6, 135.0,

131.7, 131.0, 128.3, 126.0, 125.6, 121.4, 119.5, 91.2. IR (KBr): (cm^-1): 3396, 3207, 1626,

1594, 1529, 1498, 1446, 1310, 1261. HRMS (ESI-TOF): calcd. for C₁₉H₁₄Br₂N₅O: 485.9560

[M+H]⁺, found: 485.9564.

4.5.12. 4-amino-6-phenyl-2-(phenylamino)pyrido[2,3-d]pyrimidin-7(8H)-one (15{4})

As above for 15{2} but using 4-amino-6-phenyl-2-(phenylamino)-5,8-dihydropyrido[2,3-

d]pyrimidin-7(6H)-one (14{4}) (165.7 mg, 0.5 mmol), sodium hydride (NaH) (60%

dispersion in mineral oil) (60.0 mg, 1.5 mmol) and anhydrous DMSO (5 mL) to afford 157.0

mg (0.48 mmol, 95%) of 4-amino-6-phenyl-2-(phenylamino)pyrido[2,3-d]pyrimidin-7(8H)-

o

1

one (15{4}) as a brownish solid, mp > 280 C. H NMR (400 MHz, DMSO-d₆) δ (ppm):

11.76 (br s, 1H), 9.22 (br s, 1H), 8.29 (s, 1H), 7.91 (dd, J = 8.7, 1.0 Hz, 2H), 7.75 (dd, J =

8.3, 1.3 Hz, 2H), 7.43 – 7.33 (m, 4H), 7.30 (tt, J = 7.4, 1.3 Hz, 1H), 7.27– 7.21 (m, 2H), 6.94

13

(tt, J = 7.3, 1.1 Hz, 1H). C NMR (100.6 MHz, DMSO-d₆) δ (ppm): 162.7, 161.1, 159.2,

155.3, 140.7, 136.6, 133.0, 128.3, 128.3, 127.8, 126.9, 124.1, 121.3, 119.4, 92.0. IR (KBr):

(cm^-1): 3401, 3057, 2922, 1633, 1595, 1564, 1531, 1500, 1472, 1451, 1440, 1315, 1265,

899, 794, 752, 694. MS (ESI-TOF): m/z (%) = 330.1 (100) [M+H]⁺, 313.1 (10) [M-NH₂]⁺,

18

237.1 (1) [M-NHPh]⁺. HRMS (ESI-TOF): calcd. for C₁₉H₁₆N₅O: 330.1349 [M+H]⁺, found:

330.1349.

4.5.13.

4-amino-6-(2-fluoro-6-(trifluoromethyl)phenyl)-2-(phenylamino)pyrido[2,3-

d]pyrimidin-7(8H)-one (15{10})

A mixture of 208.7 mg (0.5 mmol) of 4-amino-6-(2-fluoro-6-(trifluoromethyl)phenyl)-2-

(phenylamino)-5,8-dihydropyrido[2,3-d]pyrimidin-7(6H)-one (14{10}) and 114.0 mg (1.0

mmol) of activated MnO₂in 8.5 mL of acetic acid was refluxed for 3 hours. The resulting hot

suspension was filtered and the solvent was removed in vacuo. The resulting solid was

refluxed with water in order to eliminate MnO₂traces, then filtrated and dried in vacuo over

phosphorus pentoxide to afford 187.3 mg (0.45 mmol, 90%) of 4-amino-6-(2-fluoro-6-

(trifluoromethyl)phenyl)-2-(phenylamino)pyrido[2,3-d]pyrimidin-7(8H)-one (15{10}) as a

brownish solid, mp > 280 ^oC. ¹H NMR (400 MHz, DMSO-d₆) δ (ppm): 11.85 (br s, 1H), 9.26

(br s, 1H), 8.07 (s, 1H), 7.89 (d, J = 8.3 Hz, 2H), 7.65 (m, 4H), 7.32 (br s, 2H), 7.25 (t, J = 8.0

13

Hz, 2H), 6.94 (t, J = 7.2 Hz, 1H). C NMR (100.6 MHz, DMSO-d₆) δ (ppm): 161.8, 161.0,

160.5 (d, J = 244.3 Hz), 159.5, 156.2, 140.5, 135.4, 130.5 (d, J = 8.9 Hz), 130.4 (dq, J = 31.2,

3.0 Hz), 128.3, 123.9 (d, J = 20.5 Hz), 123.3 (dq, J = 274.5, 2.8 Hz), 121.9 (m), 121.4, 119.7,

119.5, 116.3, 91.0. IR (KBr): (cm^-1): 3407, 3212, 1631, 1596, 1565, 1533, 1500, 1467, 1445,

1320, 1253, 1170, 1132, 902, 800. HRMS (ESI-TOF): calcd. for C₂₀H₁₄F₄N₅O: 416.1129

[M+H]⁺, found: 416.1133.

4.5.14.

4-amino-6-(2,6-dibromophenyl)-8-methyl-2-(phenylamino)pyrido[2,3-

d]pyrimidin-7(8H)-one (6{2})

To a solution of 341.0 mg (0.7 mmol) of 4-amino-6-(2,6-dibromophenyl)-2-

(phenylamino)pyrido[2,3-d]pyrimidin-7(8H)-one (15{2}) in 10 mL of anhydrous DMSO,

28.0 mg (0.7 mmol) of sodium hydride (NaH) (60% dispersion in mineral oil) were added

and the mixture was stirred for 1 hour at room temperature under nitrogen atmosphere. After

this period, 43.8 µL (0.7 mmol) of methyl iodide were added dropwise and then stirred

overnight at room temperature. The reaction was quenched by addition of 300 mL of water

and the resulting precipitate was filtered, washed with water and dried in vacuo over

phosphorus pentoxide to afford 325.0 mg (0.65 mmol, 93%) of 4-amino-6-(2,6-

dibromophenyl)-8-methyl-2-(phenylamino)pyrido[2,3-d]pyrimidin-7(8H)-one (6{2}) as a

o

1

brownish solid, mp 259-261 C. H NMR (400 MHz, DMSO-d₆) δ (ppm): 9.43 (br s, 1H),

8.05 (s, 1H), 7.84 (d, J = 8.5 Hz, 2H), 7.77 (dd, J = 8.0, 0.7 Hz, 2H), 7.40 (br s, 2H), 7.35 –

13

7.23 (m, 3H), 7.01 – 6.92 (m, 1H), 3.60 (s, 3H). C NMR (100.6 MHz, DMSO-d₆) δ (ppm):

161.7, 160.3, 159.0, 156.0, 140.4, 138.9, 133.6, 131.7, 131.0, 128.4, 125.6, 124.6, 121.7,

119.6, 91.6, 28.4. IR (KBr): (cm^-1): 3402, 2924, 1631, 1572, 1523, 1496, 1468, 1445, 1344,

1315, 1193, 799. HRMS (ESI-TOF): calcd. for C₂₀H₁₆Br₂N₅O: 499.9716 [M+H]⁺, found:

499.9717.

4.5.15. 4-amino-8-methyl-6-phenyl-2-(phenylamino)pyrido[2,3-d]pyrimidine-7(8H)-one

(6{4})

As above for 6{2} but using 4-amino-6-phenyl-2-(phenylamino)pyrido[2,3-d]pyrimidin-

7(8H)-one (15{4}) (164.7 mg, 0.5 mmol), sodium hydride (NaH) (60% dispersion in mineral

oil) (20.0 mg, 0.5 mmol), anhydrous DMSO (7 mL) and methyl iodide (31.3 µL, 0.5 mmol to

afford

143.7

mg

(0.42

mmol,

84%)

of

4-amino-8-methyl-6-phenyl-2-

(phenylamino)pyrido[2,3-d]pyrimidine-7(8H)-one (6{4}) as a brownish solid, mp 238-241

1

^oC. H NMR (400 MHz, DMSO-d₆) δ (ppm): 9.36 (br s, 1H), 8.30 (s, 1H), 7.84 (d, J = 7.7

Hz, 2H), 7.72 (dd, J = 8.5, 1.2 Hz, 2H), 7.44 (br s, 2H), 7.41 (t, J = 7.6 Hz, 2H), 7.34 – 7.24

13

(m, 3H), 6.96 (t, J = 7.3 Hz, 1H), 3.63 (s, 3H). C NMR (100.6 MHz, DMSO-d₆) δ (ppm):

19

161.9, 161.7, 158.8, 155.4, 140.5, 137.1, 131.5, 128.6, 128.5, 127.9, 127.0, 123.0, 121.6,

119.5, 92.4, 28.5. IR (KBr): (cm^-1): 3347, 3210, 3055, 2924, 1632, 1575, 1525, 1495, 1438,

1400, 1346, 1310, 1234, 1198, 1013, 797, 751, 696. HRMS (ESI-TOF): calcd for C₂₀H₁₈N₅O:

344.1506, [M+H]⁺, found: 344.1508.

4.5.16.

4-amino-6-(2-fluoro-6-(trifluoromethyl)phenyl)-8-methyl-2-

(phenylamino)pyrido[2,3-d]pyrimidine-7(8H)-one (6{10})

As above for 6{2} but using 4-amino-6-(2-fluoro-6-(trifluoromethyl)phenyl)-2-

(phenylamino)pyrido[2,3-d]pyrimidin-7(8H)-one (15{10}) (290.7 mg, 0.7 mmol), sodium

hydride (NaH) (60% dispersion in mineral oil) (28.0 mg, 0.7 mmol), anhydrous DMSO (10

mL) and methyl iodide (43.8 µL, 0.7 mmol to afford 283.2 mg (0.66 mmol, 95%) of 4-amino-

6-(2-fluoro-6-(trifluoromethyl)phenyl)-8-methyl-2-(phenylamino)pyrido[2,3-d]pyrimidine-

o

1

7(8H)-one (6{10}) as a brownish solid, mp 260-261 C. H NMR (400 MHz, acetone-d₆) δ

(ppm): 8.59 (br s, 1H), 7.99 (s, 1H), 7.91 (dd, J = 8.7, 1.0 Hz, 2H), 7.68 – 7.64 (m, 2H), 7.55

13

– 7.47 (m, 1H), 7.36 – 7.29 (m, 2H), 7.04 – 6.99 (m, 1H), 6.88 (br s 2H), 3.67 (s, 3H). C

NMR (100.6 MHz, acetone-d₆) δ (ppm): 163.0, 162.2, 162.1 (d, J = 245.5 Hz), 160.46,

157.73, 141.32, 134.02, 132.2 (dq, J = 29.8, 3.1 Hz), 131.1 (d, J = 9.0 Hz), 129.37, 125.6 (dq,

J = 20.7, 1.9 Hz), 124.6 (qd, J = 273.8, 3.6 Hz), 122.9, 122.6 (m), 120.6, 120.1 (dd, J = 23.4,

0.7 Hz), 117.4, 92.5, 28.8. IR (KBr): (cm^-1): 3350, 3206, 1633, 1611, 1573, 1523, 1469,

1442, 1320, 1169, 1134, 901, 800. Anal. calcd for C₂₁H₁₅N₅OF₄: C: 58.74%, H: 3.52%, N:

16.31%, O: 3.73%, F: 17.70%; Found: C: 58.72%, H: 3.47%, N: 16.00%. MS (70 eV, EI):

m/z (%) = 429.1 (100) [M]⁺, 410.1 (15) [M-F]⁺, 360.1 (60) [M-CF₃]⁺.

4.6. Enzymatic assay

The kinase inhibition profile of compounds was evaluated at Proqinase

(http://www.proqinase.com) by measuring residual activity values at a concentration of 10

M of the test compound in singlicate in front of EGFR wild type using the following

protocol: The compounds were dissolved to 1 x 10^-3M stock solutions in 100% DMSO.

Subsequently, 100 L of each stock solution were transferred into wells A3-F12 of a

microtiter plate (“master plate”). Wells A1-F2 were filled with 100 L 100% DMSO as

controls. 5 x 10 L of the master plate were aliquoted into 5 copy plates, which were stored at

-20 ^oC until use. For the testing of each group of up to 8 kinases, one copy plate was used. In

the process, 90 L H₂O were added to each well of a copy plate. To minimize precipitation,

the H2O was added to each well only a few minutes before the transfer of the compound

solutions into the assay plates. The plate was shaken thoroughly, resulting in a “compound

dilution plate” with a compound concentration of 1 x 10^-04M/10% DMSO. This plate was

used for the transfer of 5 L compound solution into the assay plates. The final volume of the

assay was 50 L. All compounds were tested at 1 x 10^-05M in singlicate. The final DMSO

concentration in the reaction cocktails was 1% in all cases. The compound dilution plates

were disposed at the end of each working day.

A radiometric protein kinase assay (³³PanQinase® Activity Assay) was used for

measuring the kinase activity of the corresponding protein kinases. All kinase assays were

performed in 96-well FlashPlatesTM from Perkin Elmer (Boston, MA, USA) in a 50 L

reaction volume. The reaction cocktail was pipetted in 4 steps in the following order: 10 L

of non-radioactive ATP solution (in H₂O); 25 L of assay buffer/[-³³P]-ATP mixture; 5 L of

test sample in 10% DMSO; 10 L of enzyme/substrate mixture. The assay for all protein

kinases contained 70 mM HEPES-NaOH pH 7.5, 3 mM MgCl₂, 3 mM MnCl₂, 3 M Na-

orthovanadate, 1.2 mM DTT, ATP (variable amounts, corresponding to the apparent ATP-Km

20

of the respective kinase), [-³³P]-ATP (approx. 8 x 10⁰⁵cpm per well), protein kinase

(variable amounts), and substrate (variable amounts). The protein kinase reaction cocktails

were incubated at 30 ^oC for 60 min. The reaction was stopped with 50 L of 2% (v/v) H₃PO₄,

plates were aspirated and washed two times with 200 L 0.9% (w/v) NaCl. All assays were

performed with a BeckmanCoulter Biomek 2000/SL robotic system. Incorporation of ³³Pi

(counting of “cpm”) was determined with

a

microplate scintillation counter

(Microbeta,Wallac). All protein kinase assays were performed with a BeckmanCoulter Core

robotic system. For each kinase, the median value of the cpm of six wells of column 1 of each

assay plate was defined as “low control” (n = 6). This value reflects unspecific binding of

radioactivity to the plate in the absence of a protein kinase but in the presence of the

substrate. Additionally, for each kinase the median value of the cpm of six wells of column 2

of each assay plate was taken as the “high control”, i.e. full activity in the absence of any

inhibitor (n = 6). The difference between high and low control of each enzyme was taken as

100% activity. As part of the data evaluation the low control of each kinase was subtracted

from the high control value as well as from their corresponding “compound values”. The

residual activity (in %) for each compound well was calculated by using the following

formula:

Res. Activity (%) = 100 x [(signal of compound – low control) / (high control – low

control)]

(4)

As a parameter for assay quality, the Z’-factor⁷¹for the low and high controls of each

assay plate (n = 8) was used. ProQinase's criterion for repetition of an assay plate is a Z’-

factor below 0.4⁷². Z’-factors did not drop below 0.51, indicating an excellent assay quality.

Supporting Information. Backbone RMSD of EGFR in complex with the favorable pose of

6{1}; RSMF plots of 6{1}, 6{2}, 6{4}, and 6{10a} complexes; interactions of 6{1}, 6{2}),

6{4}), and 6{10a} with EGFR; Linear correlation coefficients between these energy contributions

and calculated binding free energies. Spectral data for new compounds 11{x}, 13{x}, 14{x},

15{x} and 6{x}.

Acknowledgments

We acknowledge the Centro de Computación de Alto Rendimiento (CeCAR) for providing

computational time. This work has been supported by the Agencia Nacional de Promoción

Científica y Tecnológica, Argentina (PICT-2011-2778 to CNC) and FOCEM-Mercosur (COF

03/11). CNC thanks Molsoft LLC for providing an academic license for the ICM program.

MJL is an Argentine National Research Council (CONICET) postdoctoral fellow.

References

1.

2.

3.

4.

Yarden, Y.; Sliwkowski, M. X. Nat. Rev. Mol. Cell Biol. 2001, 2, 127.

Jardines, L.; Weiss, M.; Fowble, B.; Greene, M. Pathobiology 1993, 61, 268.

Lupu, R.; Lippman, M. E. Breast Cancer Res. Treat. 1993, 27, 83.

Fontana, X.; Ferrari, P.; Namer, M.; Peysson, R.; Salanon, C.; Bussiere, F. Anticancer

Res. 1994, 14, 2099.

5.

Hickey, K.; Grehan, D.; Reid, I. M.; O'Briain, S.; Walsh, T. N.; Hennessy, T. P. Cancer

1994, 74, 1693.

6.

Stanton, P.; Richards, S.; Reeves, J.; Nikolic, M.; Edington, K.; Clark, L.; Robertson,

G.; Souter, D.; Mitchell, R.; Hendler, F. J.; et al. Br. J. Cancer 1994, 70, 427.

7.

8.

Sharma, S. V.; Bell, D. W.; Settleman, J.; Haber, D. A. Nat. Rev. Cancer 2007, 7, 169.

Vivanco, I.; Robins, H. I.; Rohle, D.; Campos, C.; Grommes, C.; Nghiemphu, P. L.;

Kubek, S.; Oldrini, B.; Chheda, M. G.; Yannuzzi, N.; Tao, H.; Zhu, S.; Iwanami, A.; Kuga,

21

D.; Dang, J.; Pedraza, A.; Brennan, C. W.; Heguy, A.; Liau, L. M.; Lieberman, F.; Yung, W.

K.; Gilbert, M. R.; Reardon, D. A.; Drappatz, J.; Wen, P. Y.; Lamborn, K. R.; Chang, S. M.;

Prados, M. D.; Fine, H. A.; Horvath, S.; Wu, N.; Lassman, A. B.; DeAngelis, L. M.; Yong, W.

H.; Kuhn, J. G.; Mischel, P. S.; Mehta, M. P.; Cloughesy, T. F.; Mellinghoff, I. K. Cancer

Discov. 2012, 2, 458.

9.

Normanno, N.; Bianco, C.; Strizzi, L.; Mancino, M.; Maiello, M. R.; De Luca, A.;

Caponigro, F.; Salomon, D. S. Curr. Drug Targets 2005, 6, 243.

10.

11.

12.

Harari, P. M. Endocr. Relat. Cancer 2004, 11, 689.

Messersmith, W. A.; Ahnen, D. J. N. Engl. J. Med. 2008, 359, 1834.

Wakeling, A. E.; Guy, S. P.; Woodburn, J. R.; Ashton, S. E.; Curry, B. J.; Barker, A. J.;

Gibson, K. H. Cancer Res. 2002, 62, 5749.

13. Pollack, V. A.; Savage, D. M.; Baker, D. A.; Tsaparikos, K. E.; Sloan, D. E.; Moyer, J.

D.; Barbacci, E. G.; Pustilnik, L. R.; Smolarek, T. A.; Davis, J. A.; Vaidya, M. P.; Arnold, L.

D.; Doty, J. L.; Iwata, K. K.; Morin, M. J. J. Pharmacol. Exp. Ther. 1999, 291, 739.

14.

Pao, W.; Miller, V.; Zakowski, M.; Doherty, J.; Politi, K.; Sarkaria, I.; Singh, B.;

Heelan, R.; Rusch, V.; Fulton, L.; Mardis, E.; Kupfer, D.; Wilson, R.; Kris, M.; Varmus, H.

Proc. Natl. Acad. Sci. U. S. A. 2004, 101, 13306.

15.

Rusnak, D. W.; Lackey, K.; Affleck, K.; Wood, E. R.; Alligood, K. J.; Rhodes, N.;

Keith, B. R.; Murray, D. M.; Knight, W. B.; Mullin, R. J.; Gilmer, T. M. Mol. Cancer Ther.

2001, 1, 85.

16.

Lynch, T. J.; Bell, D. W.; Sordella, R.; Gurubhagavatula, S.; Okimoto, R. A.;

Brannigan, B. W.; Harris, P. L.; Haserlat, S. M.; Supko, J. G.; Haluska, F. G.; Louis, D. N.;

Christiani, D. C.; Settleman, J.; Haber, D. A. N. Engl. J. Med. 2004, 350, 2129.

17.

Paez, J. G.; Janne, P. A.; Lee, J. C.; Tracy, S.; Greulich, H.; Gabriel, S.; Herman, P.;

Kaye, F. J.; Lindeman, N.; Boggon, T. J.; Naoki, K.; Sasaki, H.; Fujii, Y.; Eck, M. J.; Sellers,

W. R.; Johnson, B. E.; Meyerson, M. Science 2004, 304, 1497.

18.

Barkovich, K. J.; Hariono, S.; Garske, A. L.; Zhang, J.; Blair, J. A.; Fan, Q. W.;

Shokat, K. M.; Nicolaides, T.; Weiss, W. A. Cancer Discov. 2012, 2, 450.

19.

20.

21.

Park, J. H.; Lemmon, M. A. Cancer Discov. 2012, 2, 398.

Blundell, T. L.; Patel, S. Curr. Opin. Pharmacol. 2004, 4, 490.

Boschelli, D. H.; Wu, Z.; Klutchko, S. R.; Showalter, H. D.; Hamby, J. M.; Lu, G. H.;

Major, T. C.; Dahring, T. K.; Batley, B.; Panek, R. L.; Keiser, J.; Hartl, B. G.; Kraker, A. J.;

Klohs, W. D.; Roberts, B. J.; Patmore, S.; Elliott, W. L.; Steinkampf, R.; Bradford, L. A.;

Hallak, H.; Doherty, A. M. Journal of medicinal chemistry 1998, 41, 4365.

22.

Hamby, J. M.; Connolly, C. J.; Schroeder, M. C.; Winters, R. T.; Showalter, H. D.;

Panek, R. L.; Major, T. C.; Olsewski, B.; Ryan, M. J.; Dahring, T.; Lu, G. H.; Keiser, J.;

Amar, A.; Shen, C.; Kraker, A. J.; Slintak, V.; Nelson, J. M.; Fry, D. W.; Bradford, L.; Hallak,

H.; Doherty, A. M. J. Med. Chem. 1997, 40, 2296.

23.

Klutchko, S. R.; Hamby, J. M.; Boschelli, D. H.; Wu, Z.; Kraker, A. J.; Amar, A. M.;

Hartl, B. G.; Shen, C.; Klohs, W. D.; Steinkampf, R. W.; Driscoll, D. L.; Nelson, J. M.;

Elliott, W. L.; Roberts, B. J.; Stoner, C. L.; Vincent, P. W.; Dykes, D. J.; Panek, R. L.; Lu, G.

H.; Major, T. C.; Dahring, T. K.; Hallak, H.; Bradford, L. A.; Showalter, H. D.; Doherty, A.

M. J. Med. Chem. 1998, 41, 3276.

24.

Galve, I.; Puig de la Bellacasa, R.; Sanchez-Garcia, D.; Batllori, X.; Teixidó, J.;

Borrell, J. I. Mol. Divers. 2012, 16, 639.

25.

Perez-Pi, I.; Berzosa, X.; Galve, I.; Teixidó, J.; Borrell, J. I. Heterocycles 2010, 82,

581.

26.

Puig de la Bellacasa, R.; Roue, G.; Balsas, P.; Perez-Galan, P.; Teixidó, J.; Colomer,

D.; Borrell, J. I. Eur. J. Med. Chem. 2014, 86, 664.

27.

Phatak, S. S.; Stephan, C. C.; Cavasotto, C. N. Exp. Opin. Drug Discov. 2009, 4, 947.

22

28.

29.

30.

31.

32.

33.

1081.

34.

35.

Jorgensen, W. L. Acc. Chem. Res. 2009, 42, 724.

Cavasotto, C. N. Curr. Top. Med. Chem. 2011, 11, 1528.

Jorgensen, W. L. Angew. Chem. Int. Ed. Engl. 2012, 51, 11680.

Cavasotto, C. N.; Orry, A. J. Curr. Top. Med. Chem. 2007, 7, 1006.

Guimaraes, C. R.; Cardozo, M. J. Chem. Inf. Model. 2008, 48, 958.

Thompson, D. C.; Humblet, C.; Joseph-McCarthy, D. J. Chem. Inf. Model. 2008, 48,

Anisimov, V. M.; Cavasotto, C. N. J. Comput. Chem. 2011, 32, 2254.

Ferrari, A. M.; Degliesposti, G.; Sgobba, M.; Rastelli, G. Bioorg. Med. Chem. 2007,

15, 7865.

36.

37.

38.

39.

Genheden, S.; Ryde, U. J. Chem. Theory Comput. 2011, 7, 3768.

Hou, T.; Wang, J.; Li, Y.; Wang, W. J. Chem. Inf. Model. 2011, 51, 69.

Spyrakis, F.; Cavasotto, C. N. Arch. Biochem. Biophys. 2015, 583, 105.

Diaz, P.; Phatak, S. S.; Xu, J.; Fronczek, F. R.; Astruc-Diaz, F.; Thompson, C. M.;

Cavasotto, C. N.; Naguib, M. ChemMedChem 2009, 4, 1615.

40. Petrov, R. R.; Knight, L.; Chen, S. R.; Wager-Miller, J.; McDaniel, S. W.; Diaz, F.;

Barth, F.; Pan, H. L.; Mackie, K.; Cavasotto, C. N.; Diaz, P. Eur. J. Med. Chem. 2013, 69,

881.

41.

42.

Phatak, S. S.; Gatica, E. A.; Cavasotto, C. N. J. Chem. Inf. Model. 2010, 50, 2119.

Yun, C. H.; Boggon, T. J.; Li, Y.; Woo, M. S.; Greulich, H.; Meyerson, M.; Eck, M. J.

Cancer Cell 2007, 11, 217.

43.

44.

Stamos, J.; Sliwkowski, M. X.; Eigenbrot, C. J. Biol. Chem. 2002, 277, 46265.

Cavasotto, C. N.; Ortiz, M. A.; Abagyan, R. A.; Piedrafita, F. J. Bioorg. Med. Chem.

Lett. 2006, 16, 1969.

45.

Anisimov, V. M.; Ziemys, A.; Kizhake, S.; Yuan, Z.; Natarajan, A.; Cavasotto, C. N. J.

Comput. Aided Mol. Des. 2011, 25, 1071.

46.

797.

47.

48.

Rastelli, G.; Del Rio, A.; Degliesposti, G.; Sgobba, M. J. Comput. Chem. 2010, 31,

Pereira, E. G.; Moreira, M. A.; Caffarena, E. R. J. Mol. Model. 2012, 18, 4333.

Camarasa, M.; Barnils, C.; Puig de la Bellacasa, R.; Teixidó, J.; Borrell, J. I. Mol.

Divers. 2013, 17, 525.

49.

50.

51.

52.

Park, J. H.; Liu, Y.; Lemmon, M. A.; Radhakrishnan, R. Biochem. J. 2012, 448, 417.

Abagyan, R.; Totrov, M.; Kuznetsov, D. J. Comput. Chem. 1994, 15, 488.

ICM. MolSoft, LLC: La Jolla, CA, 2012.

Monti, M. C.; Casapullo, A.; Cavasotto, C. N.; Tosco, A.; Dal Piaz, F.; Ziemys, A.;

Margarucci, L.; Riccio, R. Chem.-Eur. J. 2009, 15, 1155.

53. Cavasotto, C. N.; Orry, A. J.; Murgolo, N. J.; Czarniecki, M. F.; Kocsi, S. A.; Hawes,

B. E.; O'Neill, K. A.; Hine, H.; Burton, M. S.; Voigt, J. H.; Abagyan, R. A.; Bayne, M. L.;

Monsma, F. J., Jr. J. Med. Chem. 2008, 51, 581.

54.

ChemBioChem 2007, 8, 1585.

55. Rossi, M.; Rotblat, B.; Ansell, K.; Amelio, I.; Caraglia, M.; Misso, G.; Bernassola, F.;

Cavasotto, C. N.; Knight, R. A.; Ciechanover, A.; Melino, G. Cell Death Dis. 2014, 5, e1203.

56. Li, W.; Cavasotto, C. N.; Cardozo, T.; Ha, S.; Dang, T.; Taneja, S. S.; Logan, S. K.;

Garabedian, M. J. Mol. Endocrinol. 2005, 19, 2273.

Monti, M. C.; Casapullo, A.; Cavasotto, C. N.; Napolitano, A.; Riccio, R.

57.

58.

Abagyan, R.; Argos, P. J. Mol. Biol. 1992, 225, 519.

Jorgensen, W. L.; Chandrasekhar, J.; Madura, J. D.; Impey, R. W.; Klein, M. L. J.

Chem. Phys. 1983, 79, 926.

59.

Hornak, V.; Abel, R.; Okur, A.; Strockbine, B.; Roitberg, A.; Simmerling, C. Proteins

2006, 65, 712.

23

60.

Wang, J.; Wolf, R. M.; Caldwell, J. W.; Kollman, P. A.; Case, D. A. J. Comput. Chem.

2004, 25, 1157.

61.

62.

Jakalian, A.; Jack, D. B.; Bayly, C. I. J. Comput. Chem. 2002, 23, 1623.

Case, D. A.; Darden, T. A.; Cheatham, T. E. I.; Simmerling, C. L.; Wang, J.; Duke, R.

E.; Luo, R.; Walker, R. C.; Zhang, W.; Merz, K. M.; Roberts, B.; Hayik, S.; Roitberg, A.;

Seabra, G.; Swails, J.; Götz, A. W.; Kolossváry, I.; Wong, K. F.; Paesani, F.; Vanicek, J.; Wolf,

R. M.; Liu, J.; Wu, X.; Brozell, S. R.; Steinbrecher, T.; Gohlke, H.; Cai, C.; Ye, X.; Wang, J.;

Hsieh, M.-J.; Cui, G.; Roe, D. R.; Mathews, D. H.; Seetin, M. G.; Salomon-Ferrer, R.; Sagui,

C.; Babin, V.; Luchko, T.; Gusarov, S.; Kovalenko, A.; Kollman, P. A.; University of

California: San Francisco, 2012.

63.

Phillips, J. C.; Braun, R.; Wang, W.; Gumbart, J.; Tajkhorshid, E.; Villa, E.; Chipot,

C.; Skeel, R. D.; Kale, L.; Schulten, K. J. Comput. Chem. 2005, 26, 1781.

64.

65.

Ryckaert, J.-P.; Ciccotti, G.; Berendsen, H. J. C. J. Comp. Phys. 1977, 23, 327.

Srinivasan, J.; Cheatham, T. E.; Cieplak, P.; Kollman, P. A.; Case, D. A. J. Amer.

Chem. Soc. 1998, 120, 9401.

66.

67.

68.

69.

Vorobjev, Y. N.; Hermans, J. Biophys. Chem. 1999, 78, 195.

Qiu, D.; Shenkin, P. S.; Hollinger, F. P.; Still, W. C. J. Phys. Chem. A 1997, 101, 3005.

Feig, M.; Brooks, C. L., 3rd. Curr. Opin. Struct. Biol. 2004, 14, 217.

Miller, B. R.; McGee, T. D.; Swails, J. M.; Homeyer, N.; Gohlke, H.; Roitberg, A. E.

J. Chem. Theory Comput. 2012, 8, 3314.

70. Metz, A.; Pfleger, C.; Kopitz, H.; Pfeiffer-Marek, S.; Baringhaus, K. H.; Gohlke, H. J.

Chem. Inf. Model. 2012, 52, 120.

71.

72.

Zhang, J. H.; Chung, T. D.; Oldenburg, K. R. J. Biomol. Screen. 1999, 4, 67.

Iversen, P. W.; Eastwood, B. J.; Sittampalam, G. S.; Cox, K. L. J. Biomol. Screen.

2006, 11, 247.

24

List of Captions