Novel multitarget inhibitors with antiangiogenic and immunomodulator properties

By means of docking studies, seventeen compounds T.1-T17 have been designed and evaluated as multitarget inhibitors of VEGFR-2 and PD-L1 proteins in order to overcome resistance phenomena offered by cancer. All these designed molecules display a urea moiety as a common structural feature and eight of them (T.1-T8) further contain a 1,2,3triazol moiety. The antiproliferative activity of these molecules on several tumor cell lines (HT-29, MCF-7, HeLa, A549, HL-60), on the endothelial cell line HMEC-1 and on the non-tumor cell line HEK-293 has been determined. The urea derivatives were also evaluated for their antiangiogenic properties, whereby their ability to inhibit tubulogenesis and kinase activity employing flow cytometry, ELISA, immunofluorescence and western blot techniques was measured. In addition, these techniques were also employed to investigate the immunomodulator action of the synthetic compounds on the inhibition of PD-L1 and c-Myc proteins. Compound T.2, 1(3-chlorophenyl)-3-(2-(4-(4-methoxybenzyl)-1H-1,2,3-triazol-1-yl)ethyl)urea, has shown similar results to sorafenib in both down-regulation of VEGFR-2 and inhibition of the kinase activity of this receptor. Furthermore, compound T.14, (E)-1-(4chlorophenyl)-3-(3-(4-methoxystyryl)phenyl)urea, improves the effect of T.2 as regards tube formation of endothelial cells and inhibition of VEGFR-2 tyrosine kinase activity. In addition, T.14 improves the effect of the experimental drug BMS-8 in the inhibition of PD-L1 and c-Myc proteins. Novel multitarget inhibitors with antiangiogenic and immunomodulator properties Laura Conesa-Milián, Eva Falomir,* Juan Murga,* Miguel Carda, J. Alberto Marco Depart. de Q. Inorgánica y Orgánica, Univ. Jaume I, E-12071 Castellón, Spain Depart. de Q. Orgánica, Univ. de Valencia, E-46100 Burjassot, Valencia, Spain *Corresponding authors: efalomir@uji.es, jmurga@uji.es. ABSTRACT By means of docking studies, seventeen compounds T.1-T17 have been designed and evaluated as multitarget inhibitors of VEGFR-2 and PD-L1 proteins in order to overcome resistance phenomena offered by cancer. All these designed molecules display a urea moiety as a common structural feature and eight of them (T.1-T8) further contain a 1,2,3triazol moiety. The antiproliferative activity of these molecules on several tumor cell lines (HT-29, MCF-7, HeLa, A549, HL-60), on the endothelial cell line HMEC-1 and on the non-tumor cell line HEK-293 has been determined. The urea derivatives were also evaluated for their antiangiogenic properties, whereby their ability to inhibit tubulogenesis and kinase activity employing flow cytometry, ELISA, immunofluorescence and western blot techniques was measured. In addition, these techniques were also employed to investigate the immunomodulator action of the synthetic compounds on the inhibition of PD-L1 and c-Myc proteins. Compound T.2, 1(3-chlorophenyl)-3-(2-(4-(4-methoxybenzyl)-1H-1,2,3-triazol-1-yl)ethyl)urea, has shown similar results to sorafenib in both down-regulation of VEGFR-2 and inhibition of the kinase activity of this receptor. Furthermore, compound T.14, (E)-1-(4chlorophenyl)-3-(3-(4-methoxystyryl)phenyl)urea, improves the effect of T.2 as regards tube formation of endothelial cells and inhibition of VEGFR-2 tyrosine kinase activity. In addition, T.14 improves the effect of the experimental drug BMS-8 in the inhibition of PD-L1 and c-Myc proteins.By means of docking studies, seventeen compounds T.1-T17 have been designed and evaluated as multitarget inhibitors of VEGFR-2 and PD-L1 proteins in order to overcome resistance phenomena offered by cancer. All these designed molecules display a urea moiety as a common structural feature and eight of them (T.1-T8) further contain a 1,2,3triazol moiety. The antiproliferative activity of these molecules on several tumor cell lines (HT-29, MCF-7, HeLa, A549, HL-60), on the endothelial cell line HMEC-1 and on the non-tumor cell line HEK-293 has been determined. The urea derivatives were also evaluated for their antiangiogenic properties, whereby their ability to inhibit tubulogenesis and kinase activity employing flow cytometry, ELISA, immunofluorescence and western blot techniques was measured. In addition, these techniques were also employed to investigate the immunomodulator action of the synthetic compounds on the inhibition of PD-L1 and c-Myc proteins. Compound T.2, 1(3-chlorophenyl)-3-(2-(4-(4-methoxybenzyl)-1H-1,2,3-triazol-1-yl)ethyl)urea, has shown similar results to sorafenib in both down-regulation of VEGFR-2 and inhibition of the kinase activity of this receptor. Furthermore, compound T.14, (E)-1-(4chlorophenyl)-3-(3-(4-methoxystyryl)phenyl)urea, improves the effect of T.2 as regards tube formation of endothelial cells and inhibition of VEGFR-2 tyrosine kinase activity. In addition, T.14 improves the effect of the experimental drug BMS-8 in the inhibition of PD-L1 and c-Myc proteins.


Introduction
Cancer is a complex pathological process that encompasses a large group of diseases.
There is a need therefore to find new drugs able to interfere with most of these processes in order to increase treatment options and to avoid resistance mechanisms. [1] The goal of the present work is the design, synthesis and biological evaluation of molecules with anticancer activity deemed to address two different biological targets of particular relevance in the cancer process. These targets are the tyrosin kinase VEGFR-2 (Vascular Endothelial Growth Factor Receptor 2) and the PD-L1 protein (Programmed Death-Ligand 1).
The first target is related to the angiogenesis process, the growth of new blood vessels from pre-existing vasculature, which is a critical step in tumor progression. [2] Vascular endothelial growth factors (VEGFs) regulate angiogenesis by binding to their associated receptors. Therapeutic inhibition of VEGFR-2 action is now having an impact in the clinical use for the treatment of a number of diseases. [3] The second target is related to the immunosuppressive ability of tumor cells. These overexpress negative immunologic regulators and generate a protected environment. The PD-L1 protein, which is overexpressed in cancer cells, plays a key role in the deregulation of the immune system through formation of PD-1/PD-L1 complex. [4] The present research intends to design compounds with the ability to bind to PD-L1 present in tumor cells, therefore blocking their ability to evade the immune system. Antibodies targeting the PD-1/PD-L1 immune checkpoint have achieved outstanding success in recent years. Clinically available examples include nivolumab and pidilizumab. However, their high immunogenicity and low stability have led to the research for new non-peptidic molecules. [5] It has been reported that the c-Myc oncogene has a direct role in preventing immune cells from efficiently attacking tumor cells. Indeed, c-Myc fosters tumor growth by increasing the levels of two immune checkpoint proteins, CD47 and PD-L1, which help thwart the host immune response. [6] Thus, c-Myc inactivation in tumors appears to engage the immune system to elicit cellular senescence in tumor cells and to collapse the vascular endothelial cells. [7] Sorafenib ( Figure 1) is an inhibitor of the kinase domain of VEGFR-2. It bears a N,N´diarylurea fragment that has shown, both through X-ray diffraction studies [8] and through computational studies [9], to be determinant in its interaction with the kinase domain. Docking studies have revealed that sorafenib can form 5 hydrogen bonds with the kinase domain of VEGFR-2: two with Glu885, two weak links with Cys919 and one with Asp1046. The other interactions with the kinase domain are of the hydrophobic type with the pocket formed by the amino acids Leu840, Val848, Ala866, Ile888, Leu889, Val899, Phe918, Thr916 and Leu1035. It has also been established that the strongest link is formed with the Asp1046.
Brystol-Myers Squibb reported in 2015 the discovery of the first non-peptidic molecules able to inhibit the formation of the PD-1/PD-L1 complex [10]. The mode of interaction of these compounds, indicated as BMS-8, BMS-37, BMS-202 and BMS-242 in Figure 1, was established in 2016 [11]. Using differential scanning fluorimetry (DSF) techniques, it was determined that these compounds induce thermal stabilization of PD-L1. Finally, the protein bound to the substrates was crystallized and the binding sites determined. Thus, these compounds are linked in a hydrophobic groove formed by the amino acids Tyr56, Met115, Ile116, Ala121 and Tyr123 and promote the dimerization of PD-L1 protein. Thus, it is possible to inhibit the formation of the PD-1/PD-L1 complex by means of a double pathway: the inhibitors occupy part of the area involved in the PD-1/PD-L1 interaction and, in addition, when the dimer is formed between two PD-L1 molecules, one of them has the opposite orientation to the one necessary to interact with PD-1. Consequently, the interaction between PD-1 and PD-L1 is disabled. Both the binding site of sorafenib in the kinase domain of VEGFR-2 and the binding site in PD-L1 possess a hydrophobic groove in which aromatic rings can be inserted. Our goal is the design of compounds able to interact with both binding sites and act as multitarget inhibitors.

Preliminary docking study
We have carried out a docking study using AutoDock 4.2 software [12] on the kinase domain of VEGFR-2 and on the site identified in PD-L1 in order to locate relatively simple structures that are able to interact with both sites. Several general structures capable of establishing at least three of the hydrogen bonds shown by sorafenib (see Figure 2) and of interacting with the binding site in PD-L1 have been identified by us.
These structures are characterized by having a urea system (which is essential to allow its interaction with the kinase domain) to which aromatic rings are bound in order to interact with the hydrophobic grooves present in both proteins.

Figure 2.
Identified general structures able to interact with the kinase domain of VEGFR-2 and PD-L1. The hydrogen bonds that can be formed with VEGFR-2 are indicated in red. The zones of the molecules that interact with the hydrophobic pockets of both binding sites are also indicated.
As an example, the docking obtained for three of the analyzed structures is shown in These compounds establish hydrogen bonds between the OMe group and Cys919, between the two NH groups of the urea system and Glu885 and between the O of urea and Asp1046. The three aromatic rings interact with two hydrophobic zones present in the binding site. It can be seen in Figure 3B   With the objective to verify these preliminary results obtained from the in-silico study, we envisaged the synthesis and the biological evaluation of several compounds derived from the general structures D.4-D.7 shown in Figure 4. Additionally, differential scanning fluorometry and free protein detection by ELISA were performed prior to the development of the corresponding cellular assays in order to verify the interaction between the targets of study and the designed compounds (see Supporting Information S2 and S3).

2.2.Synthetic work of triazolyl-ureas T.1-T.8
To begin with, we decided to synthesize triazoles with the general structure D.4 due to its easy synthetic accessibility. These compounds were obtained in only two synthetic steps from 2-azidoethan-1-amine 1 as indicated in Scheme 1 [13]. Thus, treatment of azidoamine 1 with carbonyl diimidazole in the presence of triethylamine generated the corresponding imidazolide which was then allowed to react with a range of anilines to afford azido-ureas 2-9. These were then converted into triazolyl-ureas T.1-T8 upon copper-catalyzed 1,3-dipolar cycloaddition [14] with 1-methoxy-4-(prop-2-yn-1yl)benzene (for the preparation of this compound see the Experimental section).
Unfortunately, all attempts to prepare the ortho-halogenated isomers failed due probably to the steric hindrance caused by the halogen in ortho position.

Cell proliferation inhibition
The ability of triazolyl-ureas T.1-T.8 to inhibit cell proliferation was established by means of their IC50 values towards the human tumor cell lines HT-29 (colon adenocarcinoma), MCF-7 (breast adenocarcinoma), HeLa (epithelioid cervix carcinoma) and A549 (pulmonary adenocarcinoma), as well as towards the endothelial cell line HMEC-1 (human microvascular endothelial cells) and the non-tumor cell line HEK-293 (human embryonic kidney cells). All the IC50 values observed were higher than 200 μM in all the cell lines tested and are not listed here.

Effect on cellular VEGFR-2
The effect of compounds T.1-T.8 on VEGFR-2 in HMEC-1 cell line was determined by flow cytometry technique. For this assay, cells were incubated for 24 hours in the presence of the corresponding compounds at 100 µM concentration. Then, cells were fixed with formaldehyde and treated with anti-VEGFR-2-alexafluor®647 to quantify membrane VEGFR-2 by flow cytometry. On the other hand, permeabilization of cells with Triton X-100 prior to fixation step, allowed the quantification of total VEGFR-2 in cells. Table 1 shows the effect of the selected derivatives on VEGFR-2 expression and distribution in HMEC-1 endothelial cell line, referred to control (DMSO, 100%). were also able to downregulate the expression of VEGFR-2 but to a lesser extent. Table 2

Inhibition of VEGFR-2 kinase activity
The interaction VEGF/VEGFR-2 causes phosphorylation of the receptor and triggers the signaling cascade that promotes phosphorylation of Erk1/2 (extracellular signalregulated kinase) and, subsequently, activation of the angiogenesis process [15].
In view of the fact that triazolyl-ureas T.1-T.8 interact with VEGFR-2 on endothelial cells, their capacity to inhibit kinase activity was also studied. Thus, HMEC-1 cells were treated with the derivatives for 24 h at 100 µM, then the cells were lysed and phospho-VEGFR-2 was quantified by ELISA analysis. Western blot analysis of the compounds that were active in inhibiting phospho-VEGFR-2 allowed the relative quantification of phospho-Erk1/2. Table 3 shows the percentage of p-VEGFR-2 and p-Erk for each compound referred to control (DMSO, 100%).   On the other hand, inhibition of VEGFR-2 kinase activity in A549 cells was also tested but in this case, none of the selected derivatives improved the effect observed for sorafenib. Table 4 shows the percentage of p-VEGFR-2 and p-Erk for each compound referred to control (DMSO, 100%).

Tube formation inhibition on endothelial cells
Antiangiogenic agents mainly affect endothelial cells. Thus, the capacity of synthetic compounds to inhibit the formation of new vasculature network formed by HMEC-1 was evaluated. Therefore, cells were seeded on top of Matrigel® and simultaneously treated with different concentrations of derivatives. Pictures were taken 24 h later in order to evaluate the tube formation inhibition effect. Table 5 shows the minimum concentration at which compounds are active in the tube formation inhibition. Only T.2, T. 3 and T.6 were tested, since they showed the highest activities in the inhibition of VEGFR-2 kinase activity (see Table 3). Data from Table 5 confirm that triazolyl-ureas are able to inhibit tube formation but at higher doses than sunitinib and sorafenib, with T.2 (m-chloro) being the most potent compound.

Effect on PD-L1 and c-Myc proteins
In order to evaluate the immunomodulator properties of the synthetic derivatives, the effect on PD-L1 and c-Myc proteins by the selected compounds was studied. Thus, compounds were added to A549 cells at 100 μM concentration and, after 24 h, cells were lysed and ELISA analysis was performed to determine the relative amount of both proteins compared to DMSO treated cells. Table 6 shows the percentage of free PD-L1 detected for each compound referred to control (DMSO). The percentage of c-Myc has been quantified only for the compounds which exerted a similar or better action than the reference compound BMS-8 in the inhibition of PD-L1. From Table 6 it can be concluded that compounds T.2 (m-chloro), T.3 (p-chloro) and T.8 (p-methyl) inhibit PD-L1 and c-Myc proteins in a similar way to BMS-8.

Cell proliferation evaluation in co-cultures
Some tumor cell lines such as A549 and HT-29 show high PD-L1 overexpression. [16] Therefore, cell proliferation of these has been studied in presence of PD-1 expressing Jurkat T cells in order to evaluate whether the observed cell proliferation inhibition is due to the blockage of PD-1/PD-L1 system. Compounds which showed good PD-L1 inhibition in ELISA test were evaluated. Thus, tumor cells were treated for 24 h with the selected compounds at 200 µM in presence of Jurkat T cells and then, living cells were counted using trypan blue and a Neubauer chamber. Figure 7 shows the inhibition of tumour cell proliferation exhibited by the derivatives due to the presence of Jurkat T cells. From data provided in Figure 7, it can be concluded that the tested compounds are able to inhibit A549 cell proliferation in co-culture but do no improve the effect exerted by BMS-8. Interestingly, in the HT-29 cell line this effect is much higher for compounds T.2 (m-chloro) and T.3 (p-chloro), improving the effect exhibited by BMS-8.

Synthesis of ureas T.9-T.17
The triazolyl-ureas that contain a chlorine atom in their structure are the compounds that display the best activities in our biological tests and these were very similar to the ones exhibited by our reference compounds sorafenib and BMS-8. Therefore, we decided  [17], which upon reaction with chloroanilines gave rise to ureas T.9-T.17.

Cell proliferation inhibition
The ability of ureas T.9-T. 17 Table 7 along with IC50 values for the reference compounds sunitinib, sorafenib and BMS-8. The synthetic ureas show antiproliferative activity in the low micromolar range in all tested cell lines, comparable to that shown by reference compounds sunitinib, sorafenib and BMS-8. The (E)-p-chloro derivative T.14 was the most active one with IC50 values at submicromolar level, and exhibited lower IC50 values in cancer cell lines than in nontumour endothelial ones. This fact allowed us to study the biological activity of this compound at a concentration that does not affect non-tumour cell lines.

Induction of apoptosis
Apoptosis represents a universal and exquisitely efficient cellular suicide pathway [18]. This phenomenon can be triggered by intrinsic signals such as genotoxic stress, or extrinsic signals such as the binding of ligands to cell surface death receptors [19].
Induction of apoptosis was studied by measuring the translocation of phosphatidylserine from the cytoplasmic to the extracellular side of the plasma membrane. Thus, A549 cells were incubated for 24 h in the presence of compounds at 100 µM, after which annexin-V binding was measured by flow cytometry. The apoptotic effect of compounds T.9-T.17 is indicated in Figure 8.

Effect on cellular VEGFR-2
The effect of ureas T.9-T.17 on VEGFR-2 in A549 tumor cell line was determined by both flow cytometry and immunofluorescence techniques. For these assays, cells were incubated for 24 hours in the presence of the corresponding compounds at 10 µM concentration. Table 8 and Figure 7 show the results obtained for each compound on free VEGFR-2 presence and distribution in A549 cell line, referred to control (DMSO). (E) derivatives T.13 (m-chloro) and T.14 (p-chloro) were the best ones in reducing the presence of VEGFR-2, therefore improving the effect exerted by sorafenib. Indeed, they were able to down-regulate VEGFR-2 to the half of the control inhibiting internalization of the target and reducing the total level of the protein to 10-30 % of the control.
Immunofluorescence assay correlated well with flow cytometry results. Figure 9 shows that both compounds T.13 (B) and T.14 (C) caused a significant reduction of membrane VEGFR-2 and compound T.14 exerted a higher inhibition of the internalization of the protein to the nucleus. The effect of (E) derivatives on VEGFR-2 presence and distribution in endothelial cells was also evaluated (see Table 9), referred to control (DMSO). Results from Table 9 indicate that the (E) derivatives did not inhibit internalization of VEGFR-2 in HMEC-1 cells but reduced the presence of the target in the membrane.
Again, T.13 and T.14 exhibited the highest activities, similar to sorafenib.

Inhibition of VEGFR-2 kinase activity
The effect on kinase activity of VEGFR-2 exerted by the (E) derivatives was studied by western blot on two cell lines, A549 and HMEC-1. In this case, cells were treated for 24 h at 10 µM concentration of the corresponding compounds. Table 10 shows the relative amount of p-VEGFR-2 and p-Erk1/2 detected on A549 when treated with each compound referred to control (DMSO).   Table 10 shows that T. 13 and T.14 reduce VEGFR-2 and Erk1/2 phosphorylation in the tumor cell line in a similar way to sorafenib. However, no effect on the kinase activity of VEGFR-2 on the endothelial cell line HMEC-1 was observed for any of the tested compounds.

Tube formation inhibition on endothelial cells
The capacity of ureas T.9-T. 17 to inhibit the formation of new vasculature network formed by HMEC-1 was evaluated. Table 11 shows the minimum concentration at which compounds are active and begin to inhibit the microtube formation. A comparison of the minimum active concentration values to IC50 values for HMEC-1 cell line (see Table 7) shows that there is a correlation between antiproliferative activity and tube formation inhibition capacity, as compounds with lower IC50 values exhibit microtube inhibition activity at lower concentrations. Moreover, it is observed that some of the tested compounds are more active than sunitinib and sorafenib, particularly (E) derivatives T.13 (m-chloro) and T.14 (p-chloro), which are 100-fold more active than sorafenib. Pictures for the inhibition of neovascularization achieved by compound T.14, at different concentrations, are displayed in Figure 11.

Effect on PD-L1 and c-Myc proteins
We also evaluated the immunomodulator properties of these derivatives in the same way as described above for the triazole derivatives. Thus, the effect on PD-L1 and c-Myc proteins by the selected compounds was studied by ELISA on A549 cells. After 24 h treatment at 100 μM concentration of compounds, cells were lysed and ELISA analysis was performed to determine the relative amount of both proteins compared to DMSO treated cells. Table 12 shows the percentage of free PD-L1 detected for each compound referred to control (DMSO). Again, the percentage of c-Myc has been quantified only for the compounds which exerted similar or better action than the reference compound BMS-8 in the inhibition of PD-L1. From Table 12 it can be deduced that compounds T.13, T.14, T.16 and T.17 are able to inhibit PD-L1 and c-Myc proteins and display a stronger effect than that exerted by BMS-8, with T.14 (p-chloro) being the most active in both targets.
Subsequently, immunofluorescence was performed for compound T.14. For this assay, cells were incubated for 24 hours at 100 µM of treatment. Then, cells were fixed, permeabilized and treated with anti-PD-L1-alexafluor®647 and anti-c-Myc-FITC. The pictures shown in Figure 12 correlate well with the results obtained by ELISA test (Table   12), since compound T.14 (C) inhibits PD-L1 and c-Myc proteins in a higher proportion than BMS-8 (B).

Cell proliferation evaluation in co-cultures
We also studied the effect of compounds T.13, T.14 and T.17, which showed good

PD-L1 inhibition, in affecting tumor cell proliferation in the presence of PD-1 expressing
Jurkat T-cells. Thus, A549 were treated for 24 h with the selected compounds at 200 µM in presence of Jurkat T cells. Then, living cells were counted using trypan blue and a Neubauer chamber. Figure 13 shows the inhibition of tumor cell proliferation exhibited by T.13, T.14 and T.17 due to the presence of Jurkat T cells. Data provided in Figure 13 shows that T.13 and T.14 were able to inhibit A549 cell proliferation in co-culture improving the effect exerted by BMS-8. Therefore, it can be affirmed that these derivatives act by means of disturbing PD-1/PD-L1 interactions.

Summary and conclusions
The biological study of triazolyl ureas T.1-T. 8

Experimental procedure for the synthesis of azido-ureas 2-9.
A solution of 2-azidoethan-1-amine 1 (1 eq) in DMF was treated with Et3N (2 eq). The solution was stirred for 10 min at rt and then CDI (2 eq) was added. The resulting mixture was stirred for 20 min at rt. Then, the corresponding aniline (2 eq) was added to the reaction mixture and this was stirred overnight at 50ºC. After this time, the solvent of the reaction mixture was evaporated under reduced pressure and the crude was purified by flash silica gel chromatography (hexane:EtOAc mixtures as eluent) to afford the desired products 2-9.

Experimental procedure for the synthesis of triazolyl-ureas T.1-T.8.
A solution of 1-methoxy-4-(prop-2-yn-1-yl)benzene (1 eq) in DMF was treated with the corresponding previously prepared azido-ureas (1.2 eq). Then, a mixture of CuSO4·5H2O (0.1 eq) and sodium ascorbate (0.1 eq) in DMF/H2O (9:1) was added to the reaction mixture which was stirred for 2 h at 60ºC. After that time, the mixture was concentrated, redissolved in EtOAc, and washed with brine repeated times. Finally, the organic solvent evaporated under reduced pressure and the crude was purified by flash silica gel (hexane:EtOAc mixtures as eluent) to afford the desired products. Detailed analytical data ate given in the Supplementary Material.

Experimental procedure for the synthesis of phosphonium salt 11.
A solution of 3-nitrobenzyl bromide D10 (4.63 mmol) and triphenylphosphine (4.63 mmol) in dry CH2Cl2 (20 mL) was stirred for 3 h at rt. After the reaction was complete, the white precipitate was filtered and sequentially washed with CH2Cl2 and hexane affording 11 as a white solid (1.56 g, 71%, m. p. 273ºC).
The resulting mixture was vigorously stirred at rt for 1 h in a flask that was protected from the light. Then, the solids were filtered over Celite to remove the Zn, and were thoroughly washed with EtOAc. The filtrate was neutralized with saturated aqueous NaHCO3 and the organic phase was separated and dried over MgSO4 anhydride. After filtration and solvent evaporation, silica gel column chromatography (hexane-EtOAc mixtures as eluent) was used to isolate the stereoisomers 13 and 14 (Z/E 6:4).

Experimental procedure for hydrogenation (Scheme 2, d).
Compound 12 (2.35 mmol) was dissolved in EtOAc (30 ml) and stirred with 10 % Pd/C (300 mg) for 2 h at rt and ambient pressure under a hydrogen atmosphere. The resulting mixture was filtered through a pad of Celite and the filtrate was evaporated under reduced pressure. Then, the residue was purified by silica gel column chromatography (hexane-EtOAc mixtures as eluent) to give compound 15 as a brownish solid (66%, m. p. 81-85ºC).
The resulting mixture was stirred in the dark for 20 min at 0ºC and then for 1 h at rt. After this time, H2O (5 mL/mmol) and HCl 1 M (2.5 mL/mmol) were added to the reaction mixture, which was then extracted with CH2Cl2 (3 x 20 mL). The organic layer was washed with brine, and then dried on anhydrous Na2SO4. Removal of volatiles under reduced pressure afforded an oily residue which was subjected to column chromatography on silica-gel (hexane-EtOAc mixtures as eluent) to yield the desired products 16-18.

Experimental procedure for the synthesis of ureas T.9-T.17.
A solution of the corresponding chloroaniline (0.5 mmol) in dry THF (4 mL/mmol) was treated with Et3N (5.4 mmol) under inert atmosphere. After stirring the mixture for 5 min, the corresponding previously prepared carbamate (0.5 mmol) was added dropwise as a solution in THF (10 mL/mmol). The resulting mixture was then stirred in the dark at 40-50 ºC for 24-72 h (TLC monitoring). After this time, CH2Cl2 (15 mL) and HCl 1M were added to the reaction mixture, which was then extracted with CH2Cl2 (2 x 10 mL).
The organic layer was washed with brine and then dried on anhydrous Na2SO4. Removal of volatiles under reduced pressure afforded an oily residue which was subjected to column chromatography on silica-gel (hexane-EtOAc mixtures as eluent) to afford the desired products. Detailed analytical data ate given in the Supplementary Material.
After that, the supernatant was discarded and replaced by 100 µL of DMSO to dissolve formazan crystals. The absorbance was then read at 540 nm by spectrophotometry. For all concentrations of compound, cell viability was expressed as the percentage of the ratio between the mean absorbance of treated cells and the mean absorbance of untreated cells.
Three independent experiments were performed, and the IC50 values (i.e., concentration half inhibiting cell proliferation) were graphically determined using GraphPad Prism 4 software.

Apoptosis assay
Apoptosis was determined by quantifying FITC-Annexin V translocation by means of flow cytometry. A549 cells were incubated with compounds for 24 h and then stained following instructions of BD Apoptosis Detection TM Kit. Analysis was performed with a BD Accuri TM C6 flow cytometer.

p-Erk quantification by western blot assay
From the lysates extracted in the previous experiment (5.3.6), 100 μg of proteins were subjected to gel electrophoresis using Bolt 4-12% Bis-Tris plus gels. Then, proteins were transferred to Hybond-P polyvinylidene difluoride (PVDF) membranes using the iBlot gel transfer system, and the resulting membranes were incubated first for 1 h at rt in blocking buffer (5% non-fat dry milk in TBS 1x containing 0.1% Tween) and subsequently overnight at 4 ºC in TBST buffer primary antibody solution (Anti-ERK1+ERK2 phosphoT202+T204). After washing, membranes were incubated with the corresponding HRP-conjugated secondary antibody in blocking buffer for 1 h at rt. Next, membranes were washed extensively and immunoreactive proteins were detected by chemiluminescence (ImageQuant LAS500).