New Estrone Oxime Derivatives: Synthesis, Cytotoxic Evaluation and Docking Studies

The interest in the introduction of the oxime group in molecules aiming to improve their biological effects is increasing. This work aimed to develop new steroidal oximes of the estrane series with potential antitumor interest. For this, several oximes were synthesized by reaction of hydroxylamine with the 17-ketone of estrone derivatives. Then, their cytotoxicity was evaluated in six cell lines. An estrogenicity assay, a cell cycle distribution analysis and a fluorescence microscopy study with Hoechst 3358 staining were performed with the most promising compound. In addition, molecular docking studies against estrogen receptor α, steroid sulfatase, 17β-hydroxysteroid dehydrogenase type 1 and β-tubulin were also accomplished. The 2-nitroestrone oxime showed higher cytotoxicity than the parent compound on MCF-7 cancer cells. Furthermore, the oximes bearing halogen groups in A-ring evidenced selectivity for HepaRG cells. Remarkably, the Δ9,11-estrone oxime was the most cytotoxic and arrested LNCaP cells in the G2/M phase. Fluorescence microscopy studies showed the presence of condensed DNA typical of prophase and condensed and fragmented nuclei characteristic of apoptosis. However, this oxime promoted the proliferation of T47-D cells. Interestingly, molecular docking studies estimated a strong interaction between Δ9,11-estrone oxime and estrogen receptor α and β-tubulin, which may account for the described effects.


Introduction
Cancer is a major public health problem and is one of the leading causes of death worldwide [1]. Therefore, over the years, medicinal chemists and other researchers have been working in the development of new drugs with antitumor activity-namely, starting from molecules that already exist in nature [2]. In this context, steroids are natural compounds that are usually involved in cell proliferation and consequently in cancer development [3]. In the 90s, several steroids having very unusual and interesting structures were isolated from marine sponges. Among these, steroidal oximes isolated by Rodriguez et al. [4] from Cinachyrella sponges showed relevant antiproliferative activity against several types of cancer cells [5]. The interesting results observed in these studies stimulated researchers to prepare different series of steroidal oximes with potential anticancer interest. Classically, the oxime functional group is usually introduced by condensation of an aldehyde or a ketone, including of steroidal origin, with hydroxylamine affording, respectively, aldoximes and ketoximes [6].
In another series of steroids, pregnenolone 20-oxime derivatives showed relevant activity as CYP450 17α and 5α-R enzyme inhibitors [19][20][21][22]. The introduction of an oxime group at C6 in androstane series was explored in the development of aromatase inhibitors [23].
Concerning antiproliferative assays, previous studies showed that the presence of oximes in the steroid scaffold originated compounds with relevant potential anticancer interest. For example, 6E-hydroxyimino cholest-4-ene derivatives isolated from marine sponges showed relevant antiproliferative activities against several types of cancer cells [4,5]. Later, Cushman et al. developed several 2ME2 analogues with cancer cell growth and tubulin polymerization inhibitory effects. This series of compounds included 2-(2 ,2 ,2 -trifluoroethoxy)-and 2-ethoxy-6-oxoestradiol as well as their corresponding oximes, which demonstrated clearly higher antiproliferative effects than the ketone analogues in several cancer cell lines, including breast, prostate and colon tumoral cells. In addition, these compounds also inhibited tubulin polymerization and had low binding affinities to ERα [17,24]. In estrane steroids, Rzheznikov and co-workers [25] synthesized 9α-hydroxy,11β-nitrooxyestrone-17-oxime and evidenced in vivo its antitumor effect against breast cancers. However, when compared with the corresponding 17-ketone analogue, the presence of the oxime group seemed to have low influence in this activity. In addition, these two compounds stimulated the tumor growth by the end of a 15-day treatment course, possibly due to their estrogenic effects [25]. Concerning substituted oximes, a large number of estrone-16-oxime ethers were synthesized and their antiproliferative effects were in vitro evaluated against several cell lines. Of these, among other compounds, interesting results were observed for 3-benzyloxy-16-propionyloximino-13α-methylestrone and 3-(4-methoxybenzyloxy)-16-methoximinoestrone as well as for the unsubstituted oximes 16-oximinoestrone and 3-sulfamoyloxy-16-oximinoestrone [15]. Sánchez-Sánchez et al. [26] also evidenced the antitumor effects of steroidal sapogenin oximes on HeLa cell lines and showed that the antiproliferative activity was 2.3-2.8 times higher than the observed with diosgenin. More recently, C20 oxime ester derivatives were prepared from 16-dehydropregnenolone acetate and showed cytotoxicity against leukemia (NB4), prostate (PC-3) and HeLa cancer cells [27].
In view of the therapeutic importance of steroidal oximes, and considering our interest in developing modified estrane derivatives as anticancer agents [7,28], the present study focuses on the synthesis and antiproliferative evaluation of new E1 derivatives bearing an oxime group at C17. Their cytotoxic activities were tested using breast (MCF-7, T47D), prostate (LNCaP), liver (HepaRG), colon (Caco-2) and normal fibroblast (NHDF) cell lines. For the most promising compounds the IC 50 was determined and then an estrogenicity assay, cell cycle analysis by flow cytometry after propidium iodide staining and fluorescence microscopy using Hoechst 3358 were performed. Molecular docking studies against the ERα, ST, 17β-HSD1 and β-tubulin were also accomplished.

Chemistry
Six steroidal oximes in estrane series were synthesized as shown in Scheme 1, five of which for the first time (compounds 5, 6, 10-12), to the best of our knowledge. All compounds were characterized by spectral analysis (IR, 1 H-and 13 C-NMR; in Supplementary Materials) and HRMS was also obtained for the new prepared steroidal oximes. All spectral data are in agreement with the presented structures. The NOH signal in 1 H-NMR appeared near 10 ppm. In 13 C-NMR spectra, the signal of C17-ketone appeared near 220 ppm and the C17-NOH near 168-172 ppm. The presence of ∆ 9,11 double bond (compounds 8 and 11) was associated to the signal of C11-H that appeared at 6.06 in the 1 H-NMR spectrum [29]. The nitration reaction was performed as described by Stubenrauch et al. [30], which was applied by these authors to obtain 2-nitroestrone 3. However, as an excess of the nitrating agent was used, 2,4-dinitroestrone 4 was also formed and the mixture of products was separated by column chromatography. The yields of these nitro-steroids were similar to the previously described ones [30]. The introduction of halogens (compounds 7 and 9) and of ∆ 9,11 double bond (compound 8) were effected using methodologies already applied by us [28]. Finally, for the preparation of oximes we selected a method involving the use of EtOH, NaOH and hydroxylamine hydrochloride [8] as this is a more selective and greener strategy than other approaches that use, for example, pyridine [25]. In fact, these methods involve more toxic reagents/solvents, are more time consuming, have complex workups and can lead to lower reaction yields [31][32][33].

Cell Proliferation Studies
All compounds were in vitro tested on MCF-7, T47D, LNCaP, HepaRG, Caco-2 and NHDF cell lines by the MTT colorimetric assay. In this context, it is important to mention that the results for non-oxime compounds 7-9 were described in our previous research work [28].
Firstly, a screening study was performed at 30 µM for a first analysis of the cytotoxic effect of these compounds ( Figure 1). This screening showed that several oximes led to a higher reduction in cell proliferation than the observed with parent compounds, which was particularly evident for compounds 5 and 11 in most cell lines. In addition, these two compounds and oxime 2 were the most cytotoxic in these experimental conditions. Furthermore, the cell lines mostly affected by all compounds were MCF-7 and HepaRG. On the other hand, only compound 11 promoted a significant reduction of LNCaP cells proliferation. Relative cell proliferation of hormone-dependent (MCF-7, T47-D, LNCaP) and hormone-independent cancer cells (HepaRG, Caco-2 and NHDF) incubated for 72 h at 30 µ M with the synthesized compounds, determined by the MTT assay, spectrophotometrically quantifying formazan at 570 nm. Data are expressed as a percentage of cell proliferation relative to the negative control, are indicated as means ± SD and are representative of at least two independent experiments. *** p < 0.001, ** p < 0.01, * p < 0.05 vs. control (the Student t-test).
After the screening, for the cases where a reduction of cell proliferation was higher than 50%, the IC50 was determined (Table 1). Generally, the estimated IC50 values were in agreement with the results observed in the screening, confirming that the most potent compounds were oximes 2, 5 and 11. Of these, the most cytotoxic was Δ 9,11 -estrone oxime (compound 11) on LNCaP cells (IC50 = 3.59 µ M). In addition, the highest selectivity index was also observed for this derivative in these cells (Table 2). However, the variability of MTT assays was higher in LNCaP cells, as we and others have experienced with this  compounds, determined by the MTT assay, spectrophotometrically quantifying formazan at 570 nm. Data are expressed as a percentage of cell proliferation relative to the negative control, are indicated as means ± SD and are representative of at least two independent experiments. *** p < 0.001, ** p < 0.01, * p < 0.05 vs. control (the Student t-test).
After the screening, for the cases where a reduction of cell proliferation was higher than 50%, the IC 50 was determined (Table 1). Generally, the estimated IC 50 values were in agreement with the results observed in the screening, confirming that the most potent compounds were oximes 2, 5 and 11. Of these, the most cytotoxic was ∆ 9,11 -estrone oxime (compound 11) on LNCaP cells (IC 50 = 3.59 µM). In addition, the highest selectivity index was also observed for this derivative in these cells (Table 2). However, the variability of MTT assays was higher in LNCaP cells, as we and others have experienced with this poorly adherent cell line, and therefore the fit was less good and the uncertainty is higher in this case.  Although the structure of E1 oxime (compound 2) is widely known [8,9], few studies concerning its biological activity have been published so far. Interestingly, our data showed a good antiproliferative activity of this compound against HepaRG cells (IC 50 = 16.94 µM). When evaluating the effect of the presence of ∆ 9,11 double bond (compound 11 vs. compound 2), it is interesting to note that the effect depends on the cell line, being observed a higher cytotoxicity for compound 11 in MCF-7, LNCaP, Caco-2 and NHDF cells. Concerning the effect of A-ring modifications in these oximes, the introduction of 2-nitro group (compound 5) allowed an improvement of the cytotoxicity when comparing with its absence (steroid 2) in MCF-7, HepaRG and Caco-2 cells. In the other hand, the iodination and bromination led to a lower cytotoxicity than the non-functionalized A-ring (compound 2) and 2-nitroestrone oxime 5. However, similarly to our previously published results with 2,4-diiodo-and 2,4-dibromoestrone [28], interesting IC 50 values were observed for A-ring halogenated E1 oximes 10 and 12 on HepaRG cancer cells. In this context, it is important to mention that the nitro group(s) are susceptible to reduction by nitro reductases and the ∆ 9,11 bond of compound 11 is prone to oxidation by the CYP P450 family, which reinforces the importance of study its cytotoxic effect in hepatic cells [34]. On the other hand, 2,4-dinitroestrone oxime (compound 6) displayed low antiproliferative activity. Finally, considering the data on Figure 1 and Tables 1 and 2, it is clear that the majority of these compounds had higher cytotoxicity against tumoral (MCF-7, T47D, LNCaP, HepaRG and Caco-2) than non-tumoral (NHDF) cells.
To determine the potential estrogenic profile of the synthesized compound with the most relevant anti-proliferative activity (steroid 11), its cell growing effect was measured on the estrogen-sensitive breast cancer T47-D cells (ER + ) in serum-free culture medium. This proliferative/estrogenic activity was expressed as the difference between the cell proliferation (in percentage) caused by a given compound and the basal cell proliferation fixed at 100% (Figure 2) [35,36]. E2 was also tested as reference compound. Unfortunately, similarly to the observed with E2, compound 11 also stimulated the cell proliferation at 0.001 and 0.01 µM when compared with the negative control.
To determine the potential estrogenic profile of the synthesized compound with the most relevant anti-proliferative activity (steroid 11), its cell growing effect was measured on the estrogen-sensitive breast cancer T47-D cells (ER + ) in serum-free culture medium. This proliferative/estrogenic activity was expressed as the difference between the cell proliferation (in percentage) caused by a given compound and the basal cell proliferation fixed at 100% (Figure 2) [35,36]. E2 was also tested as reference compound. Unfortunately, similarly to the observed with E2, compound 11 also stimulated the cell proliferation at 0.001 and 0.01 µ M when compared with the negative control. In our previous study, it was also evidenced that ∆ 9,11 -E1 (compound 8) also stimulated the proliferation of T47-D cells [28] and therefore, its conversion into the oxime analogue did not eliminate this effect. In this context, Palomino et al., [37] using X-ray crystallography and molecular modeling studies, showed that the presence of ∆ 9,11 double bond caused a flattening of B, C and D rings and consequently reduced the binding to ER in 1/5 th in comparison with E2. Despite this, as evidenced by our results, the presence of C9 = C11 double bond did not eliminate the estrogenic effect characteristic of these compounds. In this context, other reports also showed that the presence of an oxime group did not eliminate this effect. In fact, OMATE had a stimulatory effect (0.15 g ± 0.01) on the uterine growth in ovariectomized rats, which was approximately 50% higher than that of EMATE (0.11 g ± 0.02) [8]. In addition, and as previously referred, despite that 9α-hydroxy,11β-nitrooxyestrone-17-oxime had relevant anti-breast cancer effects, this compound also stimulated the tumor growth after a 15-day treatment period [25]. This preliminary study seemed to suggest the estrogenic activity of compound 11. Further studies will be necessary to elucidate this activity (e.g., using a luciferase assay) [38].

Cell Survival, Cell Cycle Distribution Evaluation and Hoechst 33,258 Staining
The possible mechanism of action of compound 11 was studied by flow cytometry after PI staining. This assay was performed in LNCaP cells at 24 h post treatment, and 5-FU was used as the positive control. In this cell line, it was observed that compound 11 led to 11% reduction in cell viability ( Figure 3). This effect was similar to the one originated by 5-FU (12%). In addition to this flow cytometry study, cells were also observed using an optic microscope ( Figure 4) and, after 24 h of treatment with compound 11, it was possible to see small modifications in LNCaP cells, which lost their shape, becoming rounded, as it is characteristic to happen during mitosis.
The arrest of cell cycle progression is one of the strategies used to stop cancer proliferation [39]. In this context, some studies previously published showed the effect of steroidal and non-steroidal oximes in cell cycle [26,40]. As example, it was evidenced that E1-16-oxime ethers promoted the apoptotic HeLa cell death and modulated the cell cycle progression (arrest at G 1 ), leading to an increase in cellular shrinkage, nuclear condensation, membrane permeability, sub-diploid population and caspase-3 activity [15]. In addition, 16β-triazolyl-17α-estradiol 3-benzyl ethers of the 13α-E2 series led to a G 2 /M cell cycle arrest and caspases-3 and 9 activation [41] and ∆ 9,11 -E1 induced an arrest at G 0 /G 1 in HepaRG cell cycle [28]. Thus, the interference of compound 11 in cell cycle distribution was also evaluated by flow cytometry. Interestingly, it was found that the treatment with this steroid oxime (50 µM, 24 h) induced a G 2 /M cell cycle arrest of LNCaP cells ( Figure 5). Also, LNCaP cells were not able to pass through to the S and G 2 /M phases treated by compound 5-FU, which is in agreement with literature for these cells [42].

Cell Survival, Cell Cycle Distribution Evaluation and Hoechst 33,258 Staining
The possible mechanism of action of compound 11 was studied by flow cytometry after PI staining. This assay was performed in LNCaP cells at 24 h post treatment, and 5-FU was used as the positive control. In this cell line, it was observed that compound 11 led to 11% reduction in cell viability ( Figure 3). This effect was similar to the one originated by 5-FU (12%). In addition to this flow cytometry study, cells were also observed using an optic microscope ( Figure 4) and, after 24 h of treatment with compound 11, it was possible to see small modifications in LNCaP cells, which lost their shape, becoming rounded, as it is characteristic to happen during mitosis.   The arrest of cell cycle progression is one of the strategies used to stop cancer proliferation [39]. In this context, some studies previously published showed the effect of steroidal and non-steroidal oximes in cell cycle [26,40]. As example, it was evidenced that E1-16-oxime ethers promoted the apoptotic HeLa cell death and modulated the cell cycle progression (arrest at G1), leading to an increase in cellular shrinkage, nuclear condensation, membrane permeability, sub-diploid population and caspase-3 activity [15]. In addition, 16β-triazolyl-17α-estradiol 3-benzyl ethers of the 13α-E2 series led to a G2/M cell cycle arrest and caspases-3 and 9 activation [41] and Δ 9,11 -E1 induced an arrest at G0/G1 in HepaRG cell cycle [28]. Thus, the interference of compound 11 in cell cycle distribution was also evaluated by flow cytometry. Interestingly, it was found that the treatment with this steroid oxime (50 µ M, 24 h) induced a G2/M cell cycle arrest of LNCaP cells ( Figure 5). Also, LNCaP cells were not able to pass through to the S and G2/M phases treated by compound 5-FU, which is in agreement with literature for these cells [42].
Apoptosis is essential for maintaining the physiologic balance between cell death and cell growth [43]. Therefore, studies for understanding the cancer cell cycle, particularly the interplay with chromatin control, are providing opportunities for developing a new range of anti-cancer drugs [44]. In this context, using a preliminary assay, the Hoechst 33,258 fluorescent dye was used by us to analyze nuclei morphology of LNCaP cells by fluorescence microscopy after exposition to 50 µ M of compound 11 during 24 h ( Figure 6). Interestingly, it was observed the presence of condensed DNA, typical of prophase, and a small proportion of condensed and fragmented nuclei, typical of apop- E1-16-oxime ethers promoted the apoptotic HeLa cell death and m progression (arrest at G1), leading to an increase in cellular shrink tion, membrane permeability, sub-diploid population and caspas dition, 16β-triazolyl-17α-estradiol 3-benzyl ethers of the 13α-E2 s cycle arrest and caspases-3 and 9 activation [41] and Δ 9,11 -E1 induc HepaRG cell cycle [28]. Thus, the interference of compound 11 in was also evaluated by flow cytometry. Interestingly, it was found this steroid oxime (50 µ M, 24 h) induced a G2/M cell cycle arrest of Also, LNCaP cells were not able to pass through to the S and G compound 5-FU, which is in agreement with literature for these ce Apoptosis is essential for maintaining the physiologic balan and cell growth [43]. Therefore, studies for understanding the ca larly the interplay with chromatin control, are providing opportu new range of anti-cancer drugs [44]. In this context, using a Hoechst 33,258 fluorescent dye was used by us to analyze nuclei cells by fluorescence microscopy after exposition to 50 µ M of com ( Figure 6). Interestingly, it was observed the presence of conde prophase, and a small proportion of condensed and fragmented n tosis [45]. β-Tubulin is a protein that polymerize into microtubules cell movement, intracellular trafficking and mitosis [46]. Tubulinpaclitaxel and docetaxel [47], kill cancerous cells by inhibiting leading to mitotic arrest and cell death. In this context, as it was ob 11 promoted a cell cycle arrest at G2/M ( Figure 5) and also led to densed DNA typical of prophase, plus condensed and fragme apoptosis ( Figure 6), it can be speculated that it can act by interf similarly to which occurs with other steroids of the estrane series li 2ME2 [17,24,48]. In the next section, we studied the interaction β-tubulin by docking assay to try to better understand the cell cyc  Apoptosis is essential for maintaining the physiologic balance between cell death and cell growth [43]. Therefore, studies for understanding the cancer cell cycle, particularly the interplay with chromatin control, are providing opportunities for developing a new range of anti-cancer drugs [44]. In this context, using a preliminary assay, the Hoechst 33,258 fluorescent dye was used by us to analyze nuclei morphology of LNCaP cells by fluorescence microscopy after exposition to 50 µM of compound 11 during 24 h ( Figure 6). Interestingly, it was observed the presence of condensed DNA, typical of prophase, and a small proportion of condensed and fragmented nuclei, typical of apoptosis [45]. β-Tubulin is a protein that polymerize into microtubules, which are involved in cell movement, intracellular trafficking and mitosis [46]. Tubulin-binding drugs, such as paclitaxel and docetaxel [47], kill cancerous cells by inhibiting microtubule dynamics leading to mitotic arrest and cell death. In this context, as it was observed that compound 11 promoted a cell cycle arrest at G 2 /M ( Figure 5) and also led to the formation of condensed DNA typical of prophase, plus condensed and fragmented nuclei typical of apoptosis (Figure 6), it can be speculated that it can act by interference with β-tubulin, similarly to which occurs with other steroids of the estrane series like as estramustine and 2ME2 [17,24,48]. In the next section, we studied the interaction of compound 11 and β-tubulin by docking assay to try to better understand the cell cycle arrest.

Molecular Docking Studies
Taking into account the enzymes involved in steroidogenesis, and given the structural similarity of the compounds of our study and several of the above referred steroidal oximes acting by interaction with the mentioned targets, we aimed to evaluate the affinities of the steroids prepared by us and the proteins ERα, ST, 17β-HSD1 and β-tubulin. Molecular docking is a standard computational tool that has been successfully employed in drug design and discovery studies. Satisfactory docking results can be obtained when relatively small ligands with few rotatable bonds are docked towards protein binding pockets in which flexibility does not play an important role. However, for complex molecules (with many rotatable bonds and flexibility), the use of methodology involving theoretical docking and molecular dynamics techniques are important to overtake these limitations, because they allow for evaluating and selecting the best molecule generated in the molecular docking, which can affect the results [49][50][51]. The three-dimensional structural coordinates of these three protein receptors were obtained from PDB, and molecular docking was performed using the program AutoDock vina. To validate the standard docking method, simulations were carried out between crystallized ligands/drugs with the respective proteins and all control redocking simulations were able to reproduce the ligand-protein interaction geometries presented in the respective crystal structures with a RMSD ≤ 2.0 Å . Then, all compounds were docked for the referred targets, as observed in Table 3.

Molecular Docking Studies
Taking into account the enzymes involved in steroidogenesis, and given the structural similarity of the compounds of our study and several of the above referred steroidal oximes acting by interaction with the mentioned targets, we aimed to evaluate the affinities of the steroids prepared by us and the proteins ERα, ST, 17β-HSD1 and β-tubulin. Molecular docking is a standard computational tool that has been successfully employed in drug design and discovery studies. Satisfactory docking results can be obtained when relatively small ligands with few rotatable bonds are docked towards protein binding pockets in which flexibility does not play an important role. However, for complex molecules (with many rotatable bonds and flexibility), the use of methodology involving theoretical docking and molecular dynamics techniques are important to overtake these limitations, because they allow for evaluating and selecting the best molecule generated in the molecular docking, which can affect the results [49][50][51]. The three-dimensional structural coordinates of these three protein receptors were obtained from PDB, and molecular docking was performed using the program AutoDock vina. To validate the standard docking method, simulations were carried out between crystallized ligands/drugs with the respective proteins and all control redocking simulations were able to reproduce the ligand-protein interaction geometries presented in the respective crystal structures with a RMSD ≤ 2.0 Å. Then, all compounds were docked for the referred targets, as observed in Table 3. Table 3. Predicted affinity energies of compounds 1-12 calculated against the estrogen receptor α (ERα), steroid sulfatase (ST), 17β-hydroxysteroid dehydrogenase type 1 (17β-HSD1) and β-tubulin by AutodockTools with vina executable. Binding energies of co-crystalized ligand in the X-ray crystal structures were calculated by re-docking.

Compound
Lowest Energy (kcal.mol −1 ) The RMSD between re-docked ligands and the corresponding X-ray crystal structure coordinates was ≤2. b These values were described in a previous study [28].
The results observed in the docking simulations with ERα, ST and 17β-HSD1 for compounds 7-9 are presented in our previous work [28]. According to the data presented in Table 3, compound 11 can bind ERα in a lower energy than the control (E2). In addition, in Figure 7 can be observed that this compound can form two hydrogen bonds between its oxime group at C17 and Hist 524 and between the hydroxyl group at C3 and Glu 353 of ERα target. These interactions are similar to the observed with E2. Therefore, globally, these docking data seem to be in agreement with our experimental results (Figure 2).
The predicted affinity energies of the synthesized compounds for ST are all higher than the energy obtained for co-crystalized ligand (Table 3). This suggested that these compounds have a poor affinity to this macromolecule. Concerning the energy values obtained in the docking studies with 17β-HSD1, generally they were very close to the determined affinity of the co-crystalized ligand (5α-dihydrotestosterone, DHT) ( Table 3). Due to these interesting results, we also analyzed the interaction mode of the best ranked compounds, 5 and 11, with 17β-HSD1. It is already known from the literature that the main interactions between DHT and this enzyme are a conventional hydrogen bond with Hist 221 residue and Van der Waals interactions with Leu 149, Val 143 and Pro 187 residues [52]. However, the studied compounds just exhibited the Van der Waals interactions, lacking the hydrogen bond with Hist221, which can be determinant for their interaction with this target. Further in vitro studies will be necessary to elucidate the significance of this interaction. The most interesting binding was observed between compound 11 and β-tubulin, as shown in Figure 8 and Table 3. Besides the good affinity energy value, which was lower than the determined for colchicine, compound 11 was also predicted to have the most important interactions with β-tubulin, such as the conventional hydrogen bond with Cys B 241, alkyl and Pi-alkyl interactions with Leu B 248, and Van der Waals interaction with Val B 318 [53][54][55]. Interestingly, previous studies suggested that the tubulin ligand interactions through amino acid residues Ala316 and Val318 are very crucial in inducing antitubulin effect [55]. The other studied compounds, despite having good binding energies, did not establish the conventional bonds to tubulin. Compounds 8 and 9 did not show a conventional hydrogen bond with Cys B 241, interacting only through Van der Waals, alkyl and Pi-alkyl interactions, which explain that the affinity energies may not be directly related to the established interactions required with the active site. Therefore, the cell cycle arrest at G 2 /M originated by compound 11 (Figure 5), at prophase (Figure 6), can perhaps occur due to β-tubulin inhibition. However, future studies are needed to prove this hypothesis. The predicted affinity energies of the synthesized compounds for ST are all higher than the energy obtained for co-crystalized ligand (Table 3). This suggested that these compounds have a poor affinity to this macromolecule. Concerning the energy values obtained in the docking studies with 17β-HSD1, generally they were very close to the determined affinity of the co-crystalized ligand (5α-dihydrotestosterone, DHT) ( Table 3). Due to these interesting results, we also analyzed the interaction mode of the best ranked compounds, 5 and 11, with 17β-HSD1. It is already known from the literature that the main interactions between DHT and this enzyme are a conventional hydrogen bond with Hist 221 residue and Van der Waals interactions with Leu 149, Val 143 and Pro 187 residues [52]. However, the studied compounds just exhibited the Van der Waals interactions, lacking the hydrogen bond with Hist221, which can be determinant for their interaction with this target. Further in vitro studies will be necessary to elucidate the significance of this interaction. The most interesting binding was observed between compound 11 and β-tubulin, as shown in Figure 8 and Table 3. Besides the good affinity energy value, which was lower than the determined for colchicine, compound 11 was also predicted to have the most important interactions with β-tubulin, such as the conventional hydrogen bond with Cys B 241, alkyl and Pi-alkyl interactions with Leu B 248, and Van der Waals interaction with Val B 318 [53][54][55]. Interestingly, previous studies suggested that the tubulin ligand interactions through amino acid residues Ala316 and Val318 are very crucial in inducing antitubulin effect [55]. The other studied compounds, despite having good binding energies, did not establish the conventional bonds to tubulin. Compounds 8 and 9 did not show a conventional hydrogen bond with Cys B 241, interacting only through Van der Waals, alkyl and Pi-alkyl interactions, which explain that the affinity energies may not be directly related to the established interactions required with the active site. Therefore, the cell cycle arrest at G2/M originated by compound 11 (Figure 5), at prophase (Figure 6), can perhaps occur due to β-tubulin inhibition. However, future studies are needed to prove this hypothesis.

Chemistry
All chemicals received from suppliers were used without further purification. The following reagents were purchased from:

Chemistry
All chemicals received from suppliers were used without further purification. The following reagents were purchased from: and deuterated chloroform (CDCl 3 ) were purchased from Armar Chemicals (Leipzig, Germany). All reactions were monitored by thin layer chromatography (TLC) using Al-backed aluminum/silica gel plate 0.20 mm (Macherey-Nagel 60 F254, Duren, Germany). After elution, plates were visualized in a CN-15.LC UV chamber under ultraviolet (UV) radiation (254 nm). Then, the EtOH/concentrated sulfuric acid (95:5, v:v) solution was used, followed by heating at 120 • C, to reveal the plates. The evaporation of solvents was achieved by using a rotary vacuum drier from Büchi (R-215). Infrared (IR) spectra were collected on a Thermoscientific Nicolet iS10 equipped with a diamond attenuated total reflectance crystal at room temperature in the 4000-400 cm −1 range by averaging 16 scans at a spectral resolution of 2 cm −1 . Nuclear magnetic resonance (NMR) spectra ( 1 H-NMR and 13 C-NMR) were acquired on a Bruker Avance 400 MHz spectrometer and were processed with the software TOPSPIN 4.07 (Bruker, Fitchburg, WI, USA). Chemical shifts are reported in parts per million (ppm) relative to tetramethylsilane (TMS) or solvent as an internal standard. Coupling constants (J values) are reported in hertz (Hz) and splitting multiplicities are described as s = singlet; brs = broad singlet; d = doublet and dd = double doublet. High resolution mass spectrometry (ESI-HRMS) was performed by the microanalysis service on a QSTAR XL instrument (Salamanca, Spain).

Preparation of Stock Solutions
The stock solutions of compounds were prepared in DMSO at 10 mM and stored at 4-8 • C. From these, the different diluted solutions of compounds were prepared in the corresponding complete culture medium before each experiment. The maximum DMSO concentration in cell studies was 1%.

Antiproliferative Activities against Six Cell Lines
The antiproliferative effect of compounds was evaluated by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; Sigma-Aldrich, St Louis, MO, USA) assay in MCF-7, T47-D, LNCaP, HepaRG, Caco-2 and NHDF cells. After reaching near confluence, cells were trypsinized and counted with a hemocytometer by means of the trypan-blue exclusion of dead cells. Then, 100 µL of cell suspension (2 × 10 4 cells/mL) was seeded in 96-well culture plates and left to adhere and growth during 48 h. After this period, the medium was replaced by solutions of the compounds in study (30 µM for screening assays and 0.1, 1, 10, 25, 50 and 100 µM for concentration-response studies) in the appropriate cell culture medium for approximately 72 h. Then, cells were washed with 100 µL of phosphate buffer saline (PBS; NaCl 137 mM, KCl 2.7 mM, Na 2 HPO 4 10 mM and KH 2 PO 4 1.8 mM, pH 7.4), and 100 µL of the MTT solution (5 mg/mL), prepared in the appropriate serum-free medium, was added to each well, followed by incubation for approximately 4 h at 37 • C. Afterward, MTT containing medium was removed and formazan crystals were dissolved in DMSO. Absorbance was measured at 570 nm using a microplate reader Bio-rad Xmark spectrophotometer. After background subtraction, cell proliferation values were expressed as percentage relatively to the absorbance determined in negative control cells. Untreated cells were used as the negative control and the clinical drug 5-fluorouracil (5-FU) was used as positive control. Each experiment was performed in quadruplicate and independently repeated.

E-Screening Assay in T47-D Cells
Breast T47-D cells (2 × 10 4 cells/mL; 100 µL) were seeded in 96-well culture plates in RPMI 1640 medium supplemented with 10% FBS and allowed to attach. After overnight incubation, the medium was replaced every 3 days with fresh phenol red free RPMI 1640 medium supplemented with 5% of dextran-coated charcoal-treated fetal calf serum (DCC-FCS) and containing compound 11 (0.1, 0.01 and 0.001 µM). After 6 days of exposure, the proliferation of T47-D cells was estimated by the MTT assay as described in the previous section. Each experiment was performed in quadruplicate and independently repeated. After background subtraction, cell proliferation values were expressed as percentage relatively to the absorbance determined in negative control cells.

Analysis of LNCaP Cells Viability by Flow Cytometry
The analysis of LNCaP cells viability was performed by flow cytometry after staining with propidium iodide (PI) (solution of PI 1 mg/mL in 0.1% of sodium azide and water; Sigma Aldrich, St Louis, MO, USA). Briefly, 3 mL of cells suspension were seeded in 6-well plates (5 × 10 4 cells/mL) in complete culture medium. After 48 h they were treated with 50 µM of compound 11. Untreated cells were used as negative control and 5-FU (50 µM) was used as positive control. Each experiment was performed in duplicate and independently repeated. At the end of 24 h of incubation, the supernatant of each well was collected, cells were harvested by trypsinization and pooled with the supernatants. The resulting cell suspension was kept on ice, pelleted by centrifugation and resuspended in 400 µL of complete medium. Afterward, 395 µL of the cell suspension was transferred to a FACS tube and 5 µL of PI and 0.5 µL of ethylenediamine tetraacetic acid (EDTA, 0.123 M) were added. A minimum of 20,000 events was acquired using a BD Accuri C6 (San Jose, CA, USA) flow cytometer in the channels forward scatter (FSC), side scatter (SSC) and fluorescence channel-3 (FL3, for PI). Acquisition and analysis were performed with BD Accuri Software. In the FSC/FL3 contour plot, three regions were created, one corresponding to viable cells (R1), another to dead cells (R2) and a third to an indeterminate cell population between the other two regions (R3) excluding debris that were not considered in the analysis (data not shown). The percentage of viability is the percentage of cells in R1 as compared to the total number of events in R1, R2 and R3.

Cell Cycle Distribution of LNCaP Cells
After 24 h of treatment with compound 11 at 50 µM (6-well plates, 5 × 10 4 cells/mL), LNCaP cells were collected and washed with PBS and resuspended in 450 µL of a cold solution of 0.5% bovine serum albumin (BSA; Amresco, Solon, OH, USA) and 1 mM EDTA in PBS, followed by fixation with 70% EtOH and kept at −20 • C. Afterward, fixed cells were washed twice with PBS and resuspended in a solution of PI (50 µg/mL) prepared in 0.5% BSA and 1 mM EDTA in PBS and then incubated with Ribonuclease A from bovine pancreas at a final concentration of 0.5 µg/µL (solution in 50% glycerol, 10 mM Tris-HCl, pH 8; Sigma Aldrich, St Louis, MO, USA) for 15 min in the dark. For comparison, untreated cells were used as negative control and cells treated with 5-FU at 50 µM were used as positive control. Each experiment was performed in duplicate and independently repeated. A minimum of 10,000 events was acquired using BD Accuri Software and analysis was performed by Modfit software (Becton Dickinson, San Jose, CA, USA).

Fluorescence Microscopy in LNCaP Cells after DNA Staining
Near-confluent LNCaP cells were seeded in a 6-well plate (5 × 10 4 cells/mL). After adherence and incubation for 24h with compound 11 (50 µM), the dye Hoechst 33,258 was added to the culture medium to achieve a final concentration of 1µg/mL. The cells were incubated for 15 min at 37 • C and were then photographed by means of a Nikon Eclipse microscope equipped with a fluorescence attachment containing the appropriate optical blocks and a QCapture CCD camera. Apoptosis was revealed by nuclear changes such as chromatin condensation and nuclear fragmentation.

Data Analysis
Data were expressed as mean ± standard deviation (SD). t-Student test (two groups) and one-way ANOVA (three groups) were used followed by Bonferroni post hoc tests to determine statistically significant differences among the means. Difference between groups was considered statistically significant for a p-value lower than 0.05 (p < 0.05). The determination of IC 50 was performed by sigmoidal fitting analysis [log(inhibitor) vs. normalized response-Variable slope], considering a confidence level of 95%. The coordinates of all non-standard residues, including the co-crystalized ligand, were deleted using the software Chimera (v. 1.10.1). Then, non-polar hydrogens were merged in AutoDockTools (v. 1.5.6) and Kollman and Gasteiger partial charges were added. Lastly, the prepared structure was converted from the PDB format to PDBQT for posterior use in the docking simulations.

Preparation of Ligands
All ligands used in docking simulations were built using ChemDraw (v. 12.0) software. Energy minimization and geometry optimization of these molecules were performed in Chem3D (v. 12.0) and the obtained structures were saved as PDB file format. The process of energy minimization was applied in a range from −20 to −40 kcal.mol −1 . Then, the ligands were completely prepared for docking choosing torsions and the structures were converted into PDBQT format using the software AutoDockTools.

Docking Simulations
After the preparation of macromolecules and ligands, molecular docking simulations were performed using AutoDock vina executable [62], which uses an iterated local search global optimizer. The parameter exhaustiveness of performed experiments was defined as 15. The results of molecular docking were analyzed and visualized in Discovery Studio Visualizer program from BIOVIA software.

Validation of the Molecular Docking Performance
Scoring functions are essential for molecular docking performance. In order to validate the docking performance of AutoDock vina, the difference between the real and best-scored conformations were analyzed by re-docking ERα with E2, ST with N-acetyl-D-glucosamine, 17β-HSD1 with DHT and β-tubulin with colchicine. Low root-mean-square distance (RMSD) values (<2.0 Å) were obtained for all the four cases, which means that the docking process was reliable and validated [63].

Conclusions
Several E1 oxime derivatives were synthesized and revealed interesting effects against the proliferation of several tumor cell lines when compared with parent ketone compounds. Of these, oxime 11 showed the highest activity against LNCaP cancer cells, as well as a very relevant selectivity index. In addition, it was also demonstrated that this compound originated cell cycle arrest in G 2 /M on these cells in prophase and condensed and fragmented nuclei characteristic of apoptosis. However, in an E-screening assay this oxime also promoted the proliferation of T47-D cells. Docking studies evidenced that compound 11 also showed relevant affinities for ERα and β-tubulin, which could explain its mechanism of action and estrogenic effect. Interestingly, the oximes bearing halogens in A-ring (2,4-diiodoestrone oxime 10 and 2,4-dibromoestrone oxime 12), evidenced a selectivity for HepaRG cancer cells. Another A-ring functionalized derivative, 2-nitroestrone oxime, but not its 2,4-dinitro analogue, showed higher cytotoxicity against HepaRG and MCF-7 cancer cells. Thus, the presence of an oxime group at C17 in functionalized E1 scaffold was shown to be a good strategy to obtain new molecules with relevant anticancer effects.