Novel Formulation of Undecylenic Acid induces Tumor Cell Apoptosis

Round 1

Reviewer 1 Report

1.    Figure 1. Panel B TEM represent the comparative image of different doses, are all images shown on x50 or different?

2.    Figure 2 A B and C- the IC50 values for GS-1 (mM), and UM alone (µM) exactly similar, is it by mistake same table was inserted? Looking on the plots it is unlikely that values will be similar up to fourth digits, especially when with L-Arg other than A549 cells all cells tested are proliferating significantly, though the values are in mM verses µM but needs to be checked carefully. It is mentioned and clear with the Suppl. Figs. that combination of UA and L-Arg is not forming any new compound.

3.    It is very obvious with the apoptosis analysis that GS-1 is proapoptotic, what was the LDH levels after treatment, does the LDH release was examined? as this is very important factor for any cancer drug.

4.    MTT assays were done for the viability and toxicity, is this corelate with cell death or no growth (proliferation, DNA synthesis cell cycle analysis), or number of cells were less compared to controls/cells plated? Most of the cancer, especially some breast cancer cells regrow after remission.

5.    Figure 3E the Confocal images are shown with mitotracker for mitochondrial membrane potential but it is not clear from the image other than less fluorescence intensity, was any changes author saw like mitochondrial polarity? Or any other changes. It is clear from the Fig. 3 and 4 with lysotracker GS-1 is inducing mitophagy.

6.    Four different cancer cell lines were tested and in all IC50 is low (effective), is there any specific reason why only two cell lines were chosen for further studies?

7.    The current study is done with cell culture (in Vitro), In my opinion, in vivo studies including different cancer/tumor mouse model to the study with longer treatment regimen are warranted, to get more conclusive and clear idea about the effectiveness of anti-cancer response of GS-1, that will provide more clarity to use GS-1 for next level.

Author Response

Reviewer #1

Comment 1: Figure 1. Panel B TEM represent the comparative image of different doses, are all images shown on x50 or different?

Response: We thank the reviewer for drawing attention to this and have amended the manuscript to clarify the magnification at which the TEM images were taken. The figure legend for Figure 1 now reads as:

‘Figure 1. Biophysical characterization of GS-1. (A) Schematic representation of the structure and interactions of GS-1 components. (B-E) TEM of GS-1 was performed at (B) 0.10 mM (150,00x), (C) 0.10 mM (100,000x), (D) 1.0 mM (80,000x), and (E) 0.80 mM (200,000x).’ (Page 4, line 111-113)

Comment 2: Figure 2 A B and C- the IC50 values for GS-1 (mM), and UM alone (µM) exactly similar, is it by mistake same table was inserted? Looking on the plots it is unlikely that values will be similar up to fourth digits, especially when with L-Arg other than A549 cells all cells tested are proliferating significantly, though the values are in mM verses µM but needs to be checked carefully. It is mentioned and clear with the Suppl. Figs. that combination of UA and L-Arg is not forming any new compound.

Response: We thank the reviewer for identifying these errors in the manuscript. We have made the following changes to Figure 2 to address these errors:

• Indeed the IC50 values for UA were incorrect in the original manuscript. The GS-1 IC50 values were incorrectly copied under Figure 2B. We have now revised this and the correct UA IC50 values have now been included under Figure 2B. (Page 4, Figure 2B)
• All values are in mM. The units on Figure 2B have been changed to mM, as the initial labelling with µM was in fact an error. (Page 4, Figure 2B)

Comment 3: It is very obvious with the apoptosis analysis that GS-1 is proapoptotic, what was the LDH levels after treatment, does the LDH release was examined? as this is very important factor for any cancer drug.

Response: We agree with the reviewer’s comment that an LDH release assay would give information on whether GS-1 induces cell permeability. Instead of using an LDH release assay to assess cell permeability after GS-1 treatment, we performed flow cytometry with To-Pro-3 and Annexin V on GS­-1-treated cells to determine cell permeabilization. As To-Pro-3 can only enter cells via activated pannexin channels in apoptotic cells (To-Pro-3^intermediate) or enter completely permeabilized cells (To-Pro-3^high), data in Figure 2D and Figure 2E indicated that GS-1-treated cells were permeabilized. Given the previous characterization of this flow cytometry method for determining cell permeability (Jiang et al. (2016) Monitoring the progression of cell death and the disassembly of dying cells by flow cytometry), we decided further cell permeability assays were not required.

Comment 4: MTT assays were done for the viability and toxicity, is this corelate with cell death or no growth (proliferation, DNA synthesis cell cycle analysis), or number of cells were less compared to controls/cells plated? Most of the cancer, especially some breast cancer cells regrow after remission.

Response: The reviewer makes a good point. MTT and MTS assays correlate with cells being in a non-metabolic state, and commonly associated with cell death. We did not perform assays investigating DNA synthesis or cell cycle arrest, however we came to the conclusion that GS-1 is in fact inducing cell death based on flow cytometry analysis (Figure 2D and Figure 2E) where there was an increase in phosphatidylserine exposure (as indicated with Annexin V binding) and To-Pro-3 uptake which only occurs in dead/dying cells such as apoptotic cells. We have in addition amended the manuscript to include the following statement to clarify this to the reader:

‘…and confirm that the reduced viability observed in Figure 2A does indeed correlate with cell death.’ (Page 5, lines 143-144)

Comment 5: Figure 3E the Confocal images are shown with mitotracker for mitochondrial membrane potential but it is not clear from the image other than less fluorescence intensity, was any changes author saw like mitochondrial polarity? Or any other changes. It is clear from the Fig. 3 and 4 with lysotracker GS-1 is inducing mitophagy.

Response: We agree with the reviewer that it would be of interest to investigate the effect of GS-1 on mitochondrial polarity or other mitochondrial changes. However, we chose to only investigate the fluorescence intensity of MitoTracker Red, which, as the reviewer points out, indicated a loss of mitochondrial membrane potential upon GS-1 treatment, demonstrating a loss of mitochondrial health following GS-1 treatment and supporting the pro-apoptotic mechanism of GS-1. Identifying the loss of mitochondrial membrane potential was sufficient to suggest GS-1-inducted cytotoxicity was a cell-intrinsic, pro-apoptotic mechanism. Whilst interesting, we do not believe further characterization of mitochondrial polarity following GS-1 treatment would strengthen this conclusion, and thus these experiments fall outside the scope of this study.

To address the second part of the reviewer’s comment. The dye used in Figure 4 was BODIPY, which stains lipid droplets, and we did not use this stain to assess mitochondrial morphology or function.

Comment 6: Four different cancer cell lines were tested and in all IC50 is low (effective), is there any specific reason why only two cell lines were chosen for further studies?

Response: We selected these two cell lines (A549 and Jurkat) as we believed these to be suitable representative cell lines for further detailed studies to understand the mechanism of GS-1-induced cancer toxicity. Also, Jurkat was selected as this cell line has been well described as a model in flow cytometry methods to monitor cell permeability and death, as described in the response to comment 3 (Jiang et al. (2016) Monitoring the progression of cell death and the disassembly of dying cells by flow cytometry). A549s were chosen as a representative carcinoma cell line distinct from Jurkat. Future studies with an interest in specific cancer types could replicate these studies with GS-1 in other lines, and hence we felt that for the first report of GS-1 toxicity against cancer cells, performing these experiments with representative lines was sufficient. We have added that the following statement to clarify this to readers:

‘… with Jurkat and A549 selected as representative cell lines for further experiments’.  (Page 5, line 137-138)

Comment 7: The current study is done with cell culture (in vitro), In my opinion, in vivo studies including different cancer/tumor mouse model to the study with longer treatment regimen are warranted, to get more conclusive and clear idea about the effectiveness of anti-cancer response of GS-1, that will provide more clarity to use GS-1 for next level.

Response: We agree with the reviewer’s statement. However, in the context of the study and the 10 days allotted for revisions to be made, an in vivo model with GS-1 was not feasible given this would take months to obtain ethics approval, develop, optimize, and test. In addition, this was not suggested by the other reviewer. Whilst we completely agree that in vivo experiments with GS-1 to test efficacy would provide justification for further development, we believe this to fall outside the scope of this study of initial characterization.

Reviewer 2 Report

In this paper, to solve the solubilization problem of fatty acid drugs. author compounded undecylenic acid with arginine, and create a novel drug GS-1, then author find it can induce cytotoxicity in cancer cells via apoptosis, suggest it a promise anti-cancer agent. However, the data provided by author cannot fully support author’s conclusion. Below questions must be addressed.

1.     Figure 2A and B, author should use at least one human normal epithelial cell line as control. Because if this drug will be used for cancer therapy in the further, researchers must know that if this drug do harm to normal cell as well.

2.     Figure 2B, the IC₅₀ value of GS-1 and UA could be similar, but I do not believe they are the same. Moreover, can author double check the concentration unit of UA, it should be mM but not uM.

3.     Figure 3A and B, author must detect the level of cleaved-caspase-3 as well.

4.     Figure 3C, please provide WB and sequence evidence to prove that Caspase-3 and Caspase-7/9 had been KO in U937.

5.     Figure 3D, please provide WB and sequence evidence to prove that Bax/Bak had been KO in HCT-116.

6.     Figure 5, only use small molecule inhibitor can not fully support author’s conclusion. Author should knockdown or knockout FATP2 in cancer cell, to see whether in the absence of FATP2 affect uptake of GS-1.

7.     Figure 6, author should also knockdown DGAT1/2 here, to see whether in the absence of FATP2 affect cytotoxicity of GS-1.

Author Response

Reviewer #2

Comment: Reviewer 2 suggested that ‘moderate English revisions are required’.

Response: We thank the reviewer for the suggestion, we have accordingly reviewed the manuscript and corrected grammar throughout.

Comment 1: Figure 2A and B, author should use at least one human normal epithelial cell line as control. Because if this drug will be used for cancer therapy in the further, researchers must know that if this drug do harm to normal cell as well.

Response: We agree and thank the reviewer for this helpful suggestion. We have now performed additional viability assays on Human Umbilical Vein Endothelial Cells (HUVEC) using an MTT assay. These data demonstrate that GS-1 also has a cytotoxic effect on this normal human cell line. However, we have also now included an in vivo single ascending dose toxicology study with GS-1 that demonstrates it is safe and well tolerated in vivo. Therefore the following changes have been made to the manuscript to accommodate for this additional data:

Supplementary Figures:

• The MTT HUVEC viability assay has been added to Supplementary Figures as Supplementary Figure 4.

Results:

• The following statement has been added: ‘Further, the cytotoxicity of GS-1 was tested against the primary Human Umbilical Vein Endothelial Cells (HUVEC) and a reduction in viability was also observed, suggesting a non-tumor cell specific mode of action in vitro (Supplementary Figure 4).’ (Page 4, lines 117-119)
• The following section has been added: ‘2.3 GS-1 is well tolerated in vivo 

We also sought to determine the in vivo tolerability of GS-1 and whether there would be any toxic effects. A single ascending dose study was performed in which Sprague-Dawley Rats were subcutaneously challenged with either 10 ml/kg saline or 10 ml/kg GS-1 at 190.5 mg/kg (at 19.05 mg/ml), 381 mg/kg (at 38.1 mg/ml), or 762 mg/kg (at 76.2 mg/ml). Rats that were given 190.5 mg/kg GS-1 exhibited no signs of toxicity or change in behaviour, suggesting this dose was well tolerated in vivo. Rats that were given 381 mg/kg or 762 mg/kg GS-1 displayed abnormal behaviour and toxicities within 30 min of receiving the dose, suggesting these doses exceed safe and tolerable limits. The safe dose achieved here with 190.5 mg/kg is extremely high, and likely exceeds any therapeutic exposure. Together, these results demonstrate that GS-1 was safe and well tolerated in vivo at a dose as high as 190.5 mg/kg.‘ (Page 5, lines 148-159)

Discussion:

• The following paragraph has been added: ‘Moreover, acknowledging that GS-1 is also toxic to non-cancerous human cells, GS-1 is not tumor cell specific. However, as preliminary toxicology studies suggest that GS-1 is tolerated and safe at doses that exceed therapeutic exposure, GS-1 is still viable as a potential treatment option. Based on these studies, such high concentrations and exposure time, such as in the HUVEC viability assay, would not be physiologically relevant and unlikely to occur in in vivo settings, thereby the in vivo toxicology data is more reliable in predicting GS-1 safety profile than in vitro studies.’ (Page 13, lines 336-342)

The following statements were added to the Materials and Methods section to accommodate for the new data:

• ‘Human Umbilical Vein Endothelial cells (HUVEC) were cultured in EGM-2 (Lonza).’ (Page 16, lines 465-466)
• Section 4.5 was also modified to include the HUVEC MTT assay. (Page 16, line 494)
• ‘16 Single ascending dose toxicology study with GS-1

All animal handling and treatment was approved by the University of Melbourne Animal Ethics Committee. Fifteen male Sprague-Dawley rats were given a subcutaneous (between the shoulder blades) bolus dose of either 10 ml/kg saline or 10 ml/kg GS-1 at either 190.5 mg/kg (at 19.05 mg/ml), 381 mg/kg (at 38.1 mg/ml), or 762 mg/kg (at 76.2 mg/ml). Rats were observed for any adverse events or abnormal behaviour. Animals were euthanized after 24 h.’ (Page 21, lines 726-732)

The following ethics statement was also added to the Institutional Review Board Statement:

• The animal study protocol was approved by the Animal Ethics Committee of The University of Melbourne (AEC 1914805, approved on 1 July 2019). (Page 23, lines 756-757)

Comment 2: Figure 2B, the IC₅₀ value of GS-1 and UA could be similar, but I do not believe they are the same. Moreover, can author double check the concentration unit of UA, it should be mM but not uM.

Response: We thank the reviewer for identifying these errors in the manuscript. We have made the following changes to Figure 2 to address these errors:

• Indeed the IC50 values for UA were incorrect in the original manuscript. The GS-1 IC50 values were incorrectly copied under Figure 2B. We have now revised this and the correct UA IC50 values have now been included under Figure 2B. (Page 4, Figure 2B)
• All values are in mM. The units on Figure 2B have been changed to mM, as the initial labelling with µM was in fact an error. (Page 4, Figure 2B)

Comment 3: Figure 3A and B, author must detect the level of cleaved-caspase-3 as well.

Response: We agree with the reviewer’s comment that levels of cleaved caspase-3 must also be detected. The Jurkat blot has been re-probed with an anti-cleaved caspase-3 antibody to show levels of cleaved-caspase 3. The data does indeed demonstrate increased levels of cleaved-caspase 3 upon GS-1 treatment, hence the manuscript has been modified accordingly:

Supplementary Figures:

• The WB demonstrating cleaved-caspase 3 levels has been added to supplementary figures as Supplementary Figure 5.

Results:

• The following statement has been added: ‘Moderate levels of cleaved caspase-3 were also detected in Jurkat cells after GS-1 treatment (Supplementary Figure 5).’ (Page 5, lines 166-168)

Materials and Methods:

• The following statement was added to Section 4.8 to accommodate for the new data: ‘This process was then repeated with rabbit-anti-cleaved-caspase-3 (Cell Signalling, D175 lot #43).’ (Page 18, lines 564-565)

Comment 4: Figure 3C, please provide WB and sequence evidence to prove that Caspase-3 and Caspase-7/9 had been KO in U937.

Response: The reviewer makes a valid point. The cells used in this manuscript are caspase 3/7 KOs. The Caspase 3/7 KO U937 cells have previously been validated by the Hawkins lab (who kindly gifted them to us) and the requested data can be found in the following publication (Miles et al., 2020):

Miles, M.A. et al. (2020) “Smac mimetics can provoke lytic cell death that is neither apoptotic nor necroptotic”, Apoptosis, 25(7–8), bll 500–518. doi:10.1007/s10495-020-01610-8.

This reference has been added to the manuscript page 5, line 169.

Comment 5: Figure 3D, please provide WB and sequence evidence to prove that Bax/Bak had been KO in HCT-116.

Response: The reviewer makes a valid point. Validation of the Bax/Bak KO HCT-116 cells have previously been validated by the Puthalakath lab (who kindly gifted them to us) and the requested data can be found in the following publication (Glab et al., 2017):

Glab, J.A. et al. (2017) “DR5 and caspase-8 are dispensable in ER stress-induced apoptosis”, Cell Death and Differentiation, 24(5), bll 944–950. doi:10.1038/cdd.2017.53.

This reference has been added to the manuscript page 6, line 173.

Comment 6: Figure 5, only use small molecule inhibitor cannot fully support author’s conclusion. Author should knockdown or knockout FATP2 in cancer cell, to see whether in the absence of FATP2 affect uptake of GS-1.

Response: We agree with the reviewer’s comment that knockdown or knockout of FATP2 would indeed help to define whether the absence of FATP2 affects the uptake of GS-1 and would strengthen this finding described herein using the small molecular inhibitor Grassofermata. However, the experiments utilizing the FATP2 inhibitor used here (Grassofermata) have demonstrated a significant effect and it is reasonable to conclude that FATP2 does indeed likely regulate GS-1 uptake.  Furthermore, the short time frame given for revisions would not accommodate the time required to robustly optimize and perform these knockdown or knockout experiments. However, in accordance with the reviewer’s comment we have softened our conclusions around FATP2 involvement in GS-1 uptake; the following changes have been made to the manuscript:

Results:

• The title for section 2.6 has been modified to ‘FATP2 likely plays a role in GS-1 uptake’. (Page 9, line 218)
• The title for Figure 5 has been changed to ‘Figure 5. GS-1 likely enters cells via FATP2.’ (Page 10, line 254)

Discussion:

• ‘In addition to establishing that undecylenic acid localizes to lipid droplets, this study has also shown for the first time that undecylenic acid may utilize FATP2 to enter cells.’ (Page 15, lines 398-399)

Comment 7: Figure 6, author should also knockdown DGAT1/2 here, to see whether in the absence of FATP2 affect cytotoxicity of GS-1.

Response: We agree with the reviewer’s comment that experiments with small molecular inhibitors of DGAT1 and DGAT2 are not definitive in determining the role of lipid droplets in GS-1 toxicity, and that knockdown of DGAT1/2 would help to determine their roles in GS-1 localization and cytotoxicity. However, experiments utilizing the DGAT inhibitors, including the appropriate controls, demonstrate a considerable effect to conclude that the absence of DGAT1/2 activity does not appear to affect GS-1 cytotoxicity. Furthermore, the short time frame given for revisions would not accommodate the time required to robustly optimize and perform these knockdown or knockout experiments. However, in accordance with the reviewers comments we have softened our conclusions and the following changes have been made to the manuscript:

Results:

• The title for section 2.7 has been modified to ‘Blocking lipid droplet formation may not affect GS-1 toxicity’. (Page 11, line 263)
• The following statement has been amended to: ‘Therefore, these data together suggest that lipid droplet biogenesis and undecylenic acid sequestration to lipid droplets may not play a protective role in GS-1-induced cell death.’ (Page 11, lines 279-281)

Round 2

Reviewer 2 Report

The authors addressed my concerns properly, I do not have further question.