C-Ring Oxidized Estrone Acetate Derivatives: Assessment of Antiproliferative Activities and Docking Studies

Round 1

Reviewer 1 Report

Further studies should be conducted on higher number of synthesized compounds. Conclusions in manuscript are more like short comment on single results from performed tests. SAR studies on larger group of compounds should be performed.

Author Response

Authors’ response: Thank you very much for your comments. In a previous work, we prepared and evaluated several Δ^9,11-E1 derivatives as antiproliferative agents and demonstrated that the presence of this C-ring modification was relevant for their effects on cell proliferation [1]. Therefore, considering our continuous interest in developing antiproliferative steroids [2,3], namely belonging to the estrane series [1,4,5], in this work we synthesized C-ring oxidized derivatives of E1 acetate and analyzed their cytotoxic effects. We agree with you that a higher number of evaluated compounds should be relevant to obtain more complete SAR data. Considering this, using allylic oxidation and epoxidation reaction conditions, previously described by some of us, we attempted to synthesize other C-ring oxidized compounds from Δ^9,11-E1 acetate, but without success, as referred in the manuscript. We also would like to highlight that structural modifications on C-ring of steroids usually is difficult to perform, mainly due to steric constraints, and frequently the C-ring modified steroids are unstable. Despite this, the results herein presented showed that the presence of a 9α-OH can be of interest to develop new anticancer estrane derivatives without estrogenic activity and therefore this structural modification should be considered in future studies intending to develop new compounds with antiproliferative potential.

The conclusion section was also partially reformulated.

1. Canário, C.; Matias, M.; Brito, V. de; Santos, A.O.; Falcão, A.; Silvestre, S.; Alves, G. Δ^9,11-Estrone derivatives as potential antiproliferative agents: synthesis, in vitro biological evaluation and docking studies. Comptes rendus - Chim. 2020, 23, 201–217.
2. Salvador, J.A.R.; Carvalho, J.F.S.; Neves, M.A.C.; Silvestre, S.M.; Leitão, A.J.; Silva, M.M.C.; Sá e Melo, M.L. Anticancer steroids: linking natural and semi-synthetic compounds. Nat. Prod. Rep. 2013, 30, 324–374.
3. Brito, V.; Santos, A.O.; Almeida, P.; Silvestre, S. Novel 4-azaandrostenes as prostate cancer cell growth inhibitors: Synthesis, antiproliferative effects, and molecular docking studies. Comptes rendus - Chim. 2019, 22, 73–83.
4. Canário, C.; Silvestre, S.; Falcão, A.; Alves, G. Steroidal Oximes: Useful Compounds with Antitumor Activities. Curr. Med. Chem. 2018, 25, 660–686.
5. Canário, C.; Matias, M.; Brito, V.; Santos, A.O.; Falcão, A.; Silvestre, S.; Alves, G. New Estrone Oxime Derivatives: Synthesis, Cytotoxic Evaluation and Docking Studies. Molecules 2021, 26, 1–19.

Reviewer 2 Report

The manuscript by Canario et al. reports the synthesis of four C-ring oxidized estrone acetate derivatives and the assessment of antiproliferative activity against five cancer cell lines and docking study in Estrogen receptor a, Estrone Sulfatase, 17beta-hydroxysteroid dehydrogenase type 1.

The topic of the paper is of interest to the reader of the journal.  The research is well-structured, the text is well organized, and the findings are clearly presented. The results provided seems to be of interest, and generally call no objections.

In my opinion, this manuscript requires minor revisions that the authors should consider before finally accepting the paper for publication.

• The introduction is confusing, and the reader may benefit from the introduction of a figure.
• The authors must underly the differences in the synthetic protocol with previously published synthetic route.
• The exact value of the RMSD should be specified.
• The language should be polished, and the spelling mistakes should be corrected.

Author Response

The manuscript by Canario et al. reports the synthesis of four C-ring oxidized estrone acetate derivatives and the assessment of antiproliferative activity against five cancer cell lines and docking study in Estrogen receptor a, Estrone Sulfatase, 17beta-hydroxysteroid dehydrogenase type 1.

The topic of the paper is of interest to the reader of the journal.  The research is well-structured, the text is well organized, and the findings are clearly presented. The results provided seems to be of interest, and generally call no objections.

In my opinion, this manuscript requires minor revisions that the authors should consider before finally accepting the paper for publication.

Authors’ response: Thank you very much for your comments and useful suggestions.

The introduction is confusing, and the reader may benefit from the introduction of a figure.

Authors’ response: The introduction was reformulated and a new figure (Figure 1), which shows some described structural modifications, was included.  

The authors must underly the differences in the synthetic protocol with previously published synthetic route.

Authors’ response: The synthetic route used was similar to previously published protocols. The respective references were introduced in the “Materials and Methods” sections 2.1.1 to 2.1.4.

The exact value of the RMSD should be specified.

Authors’ response: The exact value of the RMSD for re-docking ERα with E2 was 1.03, for ST with N-acetyl-D-glucosamine was 0.98 and for 17β-HSD1 with DHT was 0.18. These values were also added in the annotation of Table 3.

The language should be polished, and the spelling mistakes should be corrected.

Authors’ response: The text was revised and changes appear in yellow in the revised manuscript.

Reviewer 3 Report

This is quiete an interesting paper. Biological properties of newly synthesized chemicals were clearly presented. The introductory story goes very well. Material and methods are presented in a good manner. Results are clear and concize. Quality of tables and figures is adequate. Discussion section is good. Conclusions are supported by the obtained data. 

Author Response

Thank you very much for your comments and for your interest in our work.

Reviewer 4 Report

Canario et al performed both in silico and wet experiments to assess the antiproliferative activities of estrogen-derived compounds and to investigate their interaction with the receptors for these compounds, in continuation with their previous works.

I have some remarks for the authors:

1) For all the assays, the untreated cells were used as the negative control. However, all compounds were dissolved in DMSO. It is true that usually, DMSO is well tolerated to the cell, however, the correct negative control should be the cells treated with an amount of DMSO equivalent to the final concentration of this compound found at the highest dose of the estrogen-derivatives used and in the same assay conditions. Could the author provide evidence that in these conditions, DMSO does not influence negatively the cell proliferation?

2) Was 5-FU diluted in DMSO as the other compounds? The authors do not explicit this in the Materials and methods section

3) Many different structures of the three steroidal targets selected by the authors for their docking studies are available in PDB. The authors should provide a rationale for the selection of the three PDB files used for docking purposes (were they selected for their higher structural quality, for the presence of similar ligands, or for other reasons?)

Results, page 8 of 20, Table 1: what do the authors mean with "ND" (not determined)? Does it mean that the cells were not tested with the compound, or that it was not possible to identify IC50 values because the compounds show no activity against these cells?

Results, page 8 of 20, Table 2: given that the authors considered compound 4 as an interesting compound, why did they not determine the SI? Moreover, the first compound should be marked as "E1" for consistency with the other tables and results. Same as Table 1 for "ND".

Results, page 12 of 20, Table 3: it is hard to distinguish among the compounds which is the one with the predicted best activity, as all the results fall into the error of energy determination of AutoDock (+1 -1.5 kcal/mol). It would be more correct to point out this in the comments.

Page 13 of 20, Figure 7 and 8: please correct "van der Wall" with "van der Waals" in the legend.

Discussion: in my opinion, docking results presented by the authors do not add significant information about the mechanism of action of their compounds, given that all the compounds bind with similar energies to the selected targets and the discussion of the interactions do not point out significant differences among the compounds. Given these results, I wonder why the authors did not perform other docking studies against other possible cell targets (if any), and I'm asking the authors to better clarify what is the relevant information associated to the docking analysis they have performed.

Discussion: On the basis of the analysis of the results and of the structure-activity relationship data, could the authors give an explanation on why compound 4 is active only in HepaRG cells and not in other model cells (unless ND in Table 1 means that the activity of the compounds was not tested in the other model cells; in this case, could the authors test compound 4 also on other model cells?)

Author Response

Canario et al performed both in silico and wet experiments to assess the antiproliferative activities of estrogen-derived compounds and to investigate their interaction with the receptors for these compounds, in continuation with their previous works.

I have some remarks for the authors:

1) For all the assays, the untreated cells were used as the negative control. However, all compounds were dissolved in DMSO. It is true that usually, DMSO is well tolerated to the cell, however, the correct negative control should be the cells treated with an amount of DMSO equivalent to the final concentration of this compound found at the highest dose of the estrogen-derivatives used and in the same assay conditions. Could the author provide evidence that in these conditions, DMSO does not influence negatively the cell proliferation?

Authors’ response: Thank you very much for your comments. Indeed, our research group has already studied the effect of the maximum concentration of DMSO used in the cell lines in study (1% in the case of the concentration of 100 µM for the performance of concentration-response studies). The results of these assays showed no significant cytotoxicity in the MTT assay for this concentration of DMSO. This information was introduced in the section “2.2.2 Stock solutions”.

2) Was 5-FU diluted in DMSO as the other compounds? The authors do not explicit this in the Materials and methods section

Authors’ response: 5-FU stock solution was also prepared in DMSO at 10 mM similarly to the synthesized compounds. This information was added in the manuscript in the section “2.2.2 Stock solutions”.

3) Many different structures of the three steroidal targets selected by the authors for their docking studies are available in PDB. The authors should provide a rationale for the selection of the three PDB files used for docking purposes (were they selected for their higher structural quality, for the presence of similar ligands, or for other reasons?)

Authors’ response: We selected these structures because the target enzymes were complexed with endogenous molecules like estradiol, estrone and DHT, which are also structurally similar to the tested compounds. In addition, the high resolution of the structure was also considered, however it was not considered as the crucial criterion of selection. An explanation of this selection was added to section 2.3.1.

Results, page 8 of 20, Table 1: what do the authors mean with "ND" (not determined)? Does it mean that the cells were not tested with the compound, or that it was not possible to identify IC50 values because the compounds show no activity against these cells?

Authors’ response: A first screening at 30 µM was carried out for all synthesized compounds on MCF-7, T47-D, LNCaP, HepaRG and Caco-2 cancer cells as well as on normal fibroblasts (NHDF) by the MTT assay. Then, IC₅₀ was determined only for compound/cell line when a reduction of cell proliferation was higher than 50 % in the initial screening. For this reason, IC₅₀ values were not determined for compounds with a reduction of cell proliferation <50% in the first screening because it was considered that these compounds have low or null interest as antiproliferative agents. The abbreviation “ND” is defined in the footnote of Table 1.

Results, page 8 of 20, Table 2: given that the authors considered compound 4 as an interesting compound, why did they not determine the SI? Moreover, the first compound should be marked as "E1" for consistency with the other tables and results. Same as Table 1 for "ND".

Authors’ response: The selectivity index is calculated by dividing the IC₅₀ value of the compound in non-tumoral cells (NHDF) by the value of its IC₅₀ on cancer cell lines. As referred in the previous answer, since compound 4 did not show a relevant antiproliferative activity in NHDF at the concentration of 30 µM (68%), the IC₅₀ value was not calculated. For this reason, the value of SI for compound 4 is not presented in Table 2. Compound E1 was uniformized in the manuscript and the abbreviation “ND” is defined in the footnote of Table 2.

Results, page 12 of 20, Table 3: it is hard to distinguish among the compounds which is the one with the predicted best activity, as all the results fall into the error of energy determination of AutoDock (+1 -1.5 kcal/mol). It would be more correct to point out this in the comments.

Authors’ response: This information was added in Table 3.

Page 13 of 20, Figure 7 and 8: please correct "van der Wall" with "van der Waals" in the legend.

Authors’ response: The legend of Figures 7 and 8 was corrected.

Discussion: in my opinion, docking results presented by the authors do not add significant information about the mechanism of action of their compounds, given that all the compounds bind with similar energies to the selected targets and the discussion of the interactions do not point out significant differences among the compounds. Given these results, I wonder why the authors did not perform other docking studies against other possible cell targets (if any), and I'm asking the authors to better clarify what is the relevant information associated to the docking analysis they have performed.

Authors’ response: Usually, the first potential targets considered to explain the possible mechanisms of steroids actions are those influencing hormonal biosynthesis (e.g. enzymes) and effects (e.g. receptors). Accordingly, in the context of the present work, estrogen receptor α, steroid sulfatase and 17β-hydroxysteroid dehydrogenase type 1 were the proteins selected by us mainly because: a) they are potential targets involved in hormone-dependent cancers like breast and prostate cancers; b) several steroidal compounds similar to those prepared and evaluated by us were described as interacting with these proteins; c) we experimentally observed selective effects of compound 5 against hormone-dependent cell lines and that, on contrary, compound 4 did not exhibit proliferative action in the E-screening assay. However, there is no doubt that the analyzed targets could not be responsible for the activity of the compounds studied and therefore additional studies would be necessary to further elucidate these activities. In this ambit, as flow cytometry assays revealed that steroid 4 induced a G0/G1 cell cycle arrest, the most logical future studies to be further considered should involve the most relevant proteins influencing the cell cycle, particularly those leading to an increase in cell percentage in G0/G1 phase and a reduction of the percentage of cells in S phase.

An explanation of this point was included at the end of section “4. Discussion”.

Discussion: On the basis of the analysis of the results and of the structure-activity relationship data, could the authors give an explanation on why compound 4 is active only in HepaRG cells and not in other model cells (unless ND in Table 1 means that the activity of the compounds was not tested in the other model cells; in this case, could the authors test compound 4 also on other model cells?)

Authors’ response: As previously referred, a screening assay was performed for all compounds at 30 μM in all cell lines and for the compound/cell line when a reduction of cell proliferation was higher than 50 %, experimental studies to determine the IC₅₀ were then carried out. Therefore, it can be stated that compound 4 has a higher activity against HepaRG cells, but it is not necessarily inactive in the other cell lines. Despite this and also considering the available experimental results, we consider that it is not possible to have a clear explanation on the fact that compound 4 is more active in HepaRG cells. However, it is important to mention that estrone and other steroids can be converted, particularly in the liver by means of CYP1A1, CYP2B6 and CYP2E1 isoenzymes in active metabolites, which reinforces the importance of study their cytotoxic effect in hepatic cells. Therefore, a possibility to be considered explaining that compound 4 is more active against HepaRG cancer cells is their potential metabolic transformation in these cells in other more active compounds. However, further experimental studies would be necessary to elucidate these activities.

Reviewer 5 Report

I have some concerns about Chemistry.

The estrone derivatives 2-6 have been reported previously. There is no need to describe synthesis of the compounds in section “Materials and methods”, since it has already been published -literature 22,24,28,29,30,32. If you have made improvements to the method of synthesis or puryfication it is enough to describe it in the “discussion”.

Author Response

Dear reviewer

Thsnk you for your comments. Indeed, we analyzed the possibility to move the chemical synthesis section to Supplementary material. However, we considered that it will be more useful for readers and researchers working in this group of compounds to maintain it in the main manuscript, because in some cases the experimental procedures were not already described in the detail as we performed and the same can said for the structural characetrization of the prepared compounds.

Reviewer 6 Report

The authors in their research article focus on the synthesis of c-ring oxidized estrone acetate derivatives and their anti-proliferative activity analysis. The authors examined the effects of the compounds on cell cycle distribution, proliferation, and pro-estrogenic effects. I have the following questions for the authors regarding their study:

Since the authors focus in-depth in their work on the activity of compound 4, the IC50 values of this compound should be presented in Table 1, alongside other examined compounds. The selectivity index should also be provided for compound 4 in Table 2.

It would be more informative to present the results of cell viability inhibition as dose-response curves rather than showing cell viability at a fixed concentration of the examined compounds.

Why was flow cytometric analysis with propidium iodide performed after a 24-h incubation and not after 72 h, as in the case of the MTT assay? After 72 h,  cell viability was distinctly reduced for compound 4 in comparison with what is presented with propidium idodide staining in Figure 3. In the description of the figure, the authors mention the analysis of three cell populations R1-R3. It should be clarified which population is considered PI-negative. Additionally, in this figure, please correct the error on the y-axis.

Since the authors performed e-screening on T47-D cells, why were subsequent experiments carried out on HepaRG cells?

Author Response

The authors in their research article focus on the synthesis of c-ring oxidized estrone acetate derivatives and their anti-proliferative activity analysis. The authors examined the effects of the compounds on cell cycle distribution, proliferation, and pro-estrogenic effects. I have the following questions for the authors regarding their study:

Since the authors focus in-depth in their work on the activity of compound 4, the IC50 values of this compound should be presented in Table 1, alongside other examined compounds. The selectivity index should also be provided for compound 4 in Table 2.

It would be more informative to present the results of cell viability inhibition as dose-response curves rather than showing cell viability at a fixed concentration of the examined compounds.

Authors’ response: Thank you very much for your comment. Firstly we performed a screening at 30 µM for all compounds on MCF-7, T47-D, LNCaP, HepaRG and Caco-2 cancer cells and on normal fibroblasts (NHDF) by the MTT assay (these results are presented in Figure 2 of this resubmitted version). Then, IC₅₀ was calculated when a reduction of cell proliferation was higher than 50 % (the determined IC₅₀ values are presented in Table 1). Therefore, the IC₅₀ for compound 4 in MCF-7, LNCaP, Caco-2 and NHDF was not calculated because in the screening, the percentage of cell proliferation was higher than 50 % as showed in the following table. Consequently, SI was also not calculated. Concerning the presentation of dose-response curves, as we are presenting the results of the screening at 30 µM for all compounds in all cell lines (Figure 2) and also all determined the IC₅₀ values (Table 1), for simplicity reasons we believe that it is not necessary to introduce all concentration-response curves in the manuscript.

Table 1. % of proliferation of compound 4 in MCF-7, T47-D, LNCaP, HepaRG and Caco-2 and NHDF by the MTT assay.

Despite all the previous information in this answer, we understand the reviewer point of view; however, given the results observed in the screening, the IC₅₀ values for compound 4 in the other cell lines, if determined, will certainly be very high and we can also expect that the SI IC₅₀(NHDF)/IC₅₀(HepaRG) will also be high. We also would like to highlight that we only focused on compound 4 in flow cytometry studies of cell cycle and analysis by the carboxyfluorescein succinimidyl ester assay after concluding that compound 5 (the most potent compound of our study) not only stimulated the proliferation of T47D cells in the E-screening assay but also lead to a marked reduction of HepaRG cells viability. Unfortunately, despite the interesting potency and selectivity observed for compound 5, due to the potential toxicity risks inherent to its potential use, we decided to not further explore its bioactivity. On the other hand, despite being clearly less potent, compound 4 did not stimulate the proliferation of T47D cells in the E-screening assay and therefore its chemical structure can be of interest in the development of new antiproliferative agents without estrogenic effect.

Why was flow cytometric analysis with propidium iodide performed after a 24-h incubation and not after 72 h, as in the case of the MTT assay? After 72 h, cell viability was distinctly reduced for compound 4 in comparison with what is presented with propidium idodide staining in Figure 3. In the description of the figure, the authors mention the analysis of three cell populations R1-R3. It should be clarified which population is considered PI-negative. Additionally, in this figure, please correct the error on the y-axis.

Authors’ response: Indeed, other experimental conditions were tested, such as the same 72 h of time used in the incubation of compounds for MTT assays. However, the optimization of the experiments showed that at 50 µM and after 24 h of exposition allowed to obtain the most reliable results. In this flow cytometry study, the population that is considered PI-negative corresponded to viable cells, and this information was added to this Figure (now, Fig.3 is Fig.4). Thank you for the mistake identification in this Figure, which was corrected.

Since the authors performed e-screening on T47-D cells, why were subsequent experiments carried out on HepaRG cells?

Authors’ response: E-screening assay using T47-D cells is a test performed to study the proliferation activity of compounds in estrogen-dependent cells, an undesirable side effect of steroids with interest, namely, in cancer treatment. Therefore, as compound 4 is not estrogenic (it did not promote a cell proliferation in our experimental conditions) and showed antiproliferative activity against liver cancer cells, the subsequent experiments only with this compound (flow cytometry studies of cell cycle, and analysis by the carboxyfluorescein succinimidyl ester assay) were performed in HepaRG cancer cells with the intention to understand its potentiality as antiproliferative agent in liver cancer.

Round 2

Reviewer 4 Report

The authors have answered to my remarks and amended, wherever necessary, the text of the manuscript accordingly. At present I can recommend acceptance of the manuscript for publication.

Reviewer 6 Report

The authors have addressed my comments and questions, thus I recommend the publication of this article.