Submit to this Journal Review for this Journal Propose a Special Issue

Article Menu

Share Help Cite Discuss in SciProfiles

Open AccessArticle

Peer-Review Record

Disulfidptosis: A New Target for Parkinson’s Disease and Cancer

Curr. Issues Mol. Biol. 2024, 46(9), 10038-10064; https://doi.org/10.3390/cimb46090600

by Tingting Liu^†

, Xiangrui Kong^† and Jianshe Wei^*

Reviewer 1: Anonymous

Reviewer 2:

Jyh Fei Liao

Reviewer 3: Anonymous

Curr. Issues Mol. Biol. 2024, 46(9), 10038-10064; https://doi.org/10.3390/cimb46090600

Submission received: 27 August 2024 / Revised: 6 September 2024 / Accepted: 10 September 2024 / Published: 12 September 2024

(This article belongs to the Section Bioinformatics and Systems Biology)

Round 1

Reviewer 1 Report (New Reviewer)

Comments and Suggestions for Authors

This study provides valuable insights into this novel cell death mechanism and its potential role in these diseases. I have reviewed the manuscript with a focus on translational aspects and future directions. While the work is scientifically sound, I have some suggestions to enhance its impact and relevance:

Translational aspects:
- Consider incorporating more human data to strengthen the clinical relevance of your findings. This could include analysis of patient samples or correlation with clinical outcomes.
- Expand the discussion on potential therapeutic implications of your findings for human patients.
- Address how the identified molecular mechanisms might vary across different human populations or disease subgroups.
Future directions:
- I strongly recommend including a section on the potential applications of artificial intelligence (AI) in advancing this field of research. Some key points to consider:
  - How AI could aid in integrating and analyzing complex multi-omics datasets
  - The potential for AI in predictive modeling of disease progression or treatment response
  - Applications in drug discovery targeting the disulfidptosis pathway
  - Use of AI for personalized medicine approaches based on molecular profiles
  - How AI might optimize experimental design, potentially reducing animal use
  - Possibilities for in silico modeling of cellular processes related to disulfidptosis
- Discuss both the potential benefits and limitations of AI applications in this context
- Emphasize a collaborative approach between AI and traditional experimental methods
Methodology:
- While your animal studies provide crucial insights, consider discussing how some aspects of the research might be complemented or partially replaced by in silico methods in the future.
Broader impact:
- Expand on how your findings on disulfidptosis might apply to other neurodegenerative diseases or cancer types beyond those directly studied.

Addressing these points would enhance the translational relevance of your work and position it within the context of emerging research technologies. This could broaden the appeal and impact of your manuscript for a diverse readership.

Author Response

Translational aspects:

Consider incorporating more human data to strengthen the clinical relevance of your findings. This could include analysis of patient samples or correlation with clinical outcomes.

Thank you very much for your careful review and valuable feedback on my research work. I strongly agree with your suggestion to enhance the clinical relevance of our research and believe that it is crucial for improving the practical value and impact of our research results.

In response to your suggestions, I plan to focus on the following aspects to better integrate human data and enhance the clinical relevance of the research:

Analyze patient samples:

I will strive to obtain patient samples from clinical partner institutions, which should cover individuals with different disease stages, treatment responses, and prognosis. By conducting in-depth analysis on these samples (such as genomics, proteomics, metabolomics, etc.), we can more directly observe the performance of my research findings in actual patients.

At the same time, I will design a rigorous experimental plan to ensure standardization and reproducibility of sample processing, data collection, and analysis processes, in order to improve the reliability and comparability of the results.

Related clinical outcomes:

I will try to work closely with the clinical team to collect clinical information from patients, including disease diagnosis, treatment plans, treatment outcomes, and long-term prognosis. These pieces of information will be used to analyze the potential association between my research findings and clinical outcomes.

Through statistical analysis and data mining techniques, I will explore whether my research findings can predict or explain clinical indicators such as patient treatment response, disease progression, or quality of life, in order to further validate their clinical value.

Expanding sample size and multi center collaboration:

In order to improve the universality and reliability of the research, I will strive to expand the sample size and seek collaboration with other medical institutions and research teams. Multi center collaboration can not only increase sample diversity and representativeness, but also improve research efficiency and impact.

Ethics and Privacy Protection:

Throughout the entire research process, I will strictly adhere to relevant ethical standards and legal regulations to ensure the confidentiality and anonymity of patient information. At the same time, I will maintain close communication with the ethics review committee to ensure that all research activities comply with ethical standards.

I am well aware that these tasks require time, resources, and multi-faceted coordination and cooperation, but I will do my best to actively promote the progress of these tasks. I believe that through these efforts, our research results will become more clinically relevant, providing stronger scientific evidence for the diagnosis, treatment, and prevention of diseases.

Thank you again for your valuable feedback. We look forward to your continued attention and guidance on our future work.

Expand the discussion on potential therapeutic implications of your findings for human patients.

Thank you very much for your careful review and valuable feedback on my research work. The identification of ACTB, ACTN4, INF2, and MYL6 as DEDRGs that are implicated in both PD and various cancer types opens up exciting possibilities for developing novel therapeutic strategies for human patients. Here, we expand on the potential therapeutic implications of our findings: Given the role of these DEDRGs in regulating cellular processes such as apoptosis, mitochondrial function, and the NF-κB signaling pathway, they represent potential therapeutic targets. For example, modulating the expression or activity of these genes through small molecules, gene therapy, or RNA interference could potentially slow disease progression or enhance the effectiveness of existing treatments. The differential expression patterns of these genes across different cancer types and subtypes suggest that precision medicine approaches could be developed. By profiling the expression of these DEDRGs in individual patients, it may be possible to tailor treatment regimens that are specifically targeted to the molecular characteristics of their disease. Our findings that these genes are associated with immune cell infiltration in cancer patients suggest that immunotherapeutic strategies may be particularly effective. Boosting the immune system's ability to target cancer cells that are expressing these DEDRGs could enhance treatment outcomes. The identification of miRNAs such as miR-4298, miR-296-3p, miR-150-3p, miR-493-5p, and miR-6742-5p as potential regulators of these DEDRGs in both PD and cancer suggests that existing drugs that target these miRNAs or their downstream pathways could be repurposed for treating these diseases. The expression levels of these DEDRGs could serve as biomarkers for early detection of PD and cancer, as well as for predicting disease progression and response to treatment. By monitoring these biomarkers in patients, clinicians could make more informed decisions about treatment options and adjust them as needed. Given the complexity of both PD and cancer, it is likely that targeting multiple pathways simultaneously will be necessary to achieve optimal therapeutic outcomes. Our findings suggest that combining therapies that target these DEDRGs with other drugs or treatments that address different aspects of the disease may yield better results. Finally, our results provide valuable insights into the molecular mechanisms underlying the relationship between PD and cancer, which could inform the development of novel drugs specifically designed to target these DEDRGs. Such drugs could potentially have broad applications in the treatment of neurodegenerative diseases and cancers. In conclusion, our study highlights the potential of ACTB, ACTN4, INF2, and MYL6 as therapeutic targets and biomarkers for PD and cancer. Further research into the roles of these genes and their downstream pathways is needed to fully realize their therapeutic potential and develop effective treatments for these devastating diseases.

Address how the identified molecular mechanisms might vary across different human populations or disease subgroups.

Thank you very much for your careful review and valuable feedback on my research work. When exploring how identified molecular mechanisms may vary in different human populations or disease subgroups, we need to conduct in-depth analysis from multiple dimensions such as genetic polymorphism, environmental interactions, disease phenotype heterogeneity, and differences in treatment response. The following are specific explanations of these aspects:

Genetic polymorphism:

There are a wide range of genetic variations in the human genome, including single nucleotide polymorphisms (SNPs), insertion/deletion variations, structural variations, etc. These variations may affect gene expression, protein structure and function, leading to differences in molecular mechanisms in different populations or disease subgroups. For example, certain gene mutations may be more common in certain populations, which may increase or decrease an individual's susceptibility to specific diseases and affect the development path and treatment outcomes of the disease.

Environmental interactions:

There are complex interactions between environmental factors (such as dietary habits, lifestyle, exposure to pollutants, etc.) and genetic background, which may affect the expression of molecular mechanisms. Different populations or disease subgroups may be exposed to different environmental factors due to geographical location, cultural customs, and other reasons, leading to variations in molecular mechanisms. For example, certain nutrients or pollutants may regulate gene expression by affecting epigenetic modifications such as DNA methylation and histone modifications, thereby affecting disease progression.

Heterogeneity of disease phenotype:

Even the same disease may exhibit significant phenotypic heterogeneity in different patients. This heterogeneity may stem from different molecular mechanisms. For example, in cancer, different patients may carry different driver gene mutations that drive different signaling pathways and cellular processes, leading to different subtypes and prognoses of the disease. Therefore, personalized treatment plans need to be developed based on the molecular mechanisms of different subgroups during treatment.

Differences in treatment response:

There may be significant differences in the response of different populations or disease subgroups to the same treatment plan. This difference can be partially attributed to differences in molecular mechanisms. For example, certain drugs may act through specific molecular targets, but if these targets have variations or differential expression in specific populations or disease subgroups, the efficacy of the drug may be affected. Therefore, genetic testing and molecular typing before treatment can help predict the patient's treatment response and optimize the treatment plan.

In summary, the variations of identified molecular mechanisms in different human populations or disease subgroups are complex, involving multiple aspects such as genetic polymorphism, environmental interactions, disease phenotype heterogeneity, and differences in treatment response. These mutations not only affect our understanding of the nature of diseases, but also pose new challenges and opportunities for disease prevention, diagnosis, and treatment. Therefore, in future research, we need to further explore the molecular mechanisms behind these variations in order to provide strong support for precision medicine of diseases.

Future directions:

I strongly recommend including a section on the potential applications of artificial intelligence (AI) in advancing this field of research. Some key points to consider:

How AI could aid in integrating and analyzing complex multi-omics datasets

The potential for AI in predictive modeling of disease progression or treatment response

Applications in drug discovery targeting the disulfidptosis pathway

Use of AI for personalized medicine approaches based on molecular profiles

How AI might optimize experimental design, potentially reducing animal use

Possibilities for in silico modeling of cellular processes related to disulfidptosis

Discuss both the potential benefits and limitations of AI applications in this context

Emphasize a collaborative approach between AI and traditional experimental methods

Thank you very much for your careful review and valuable feedback on my research work. Incorporating the potential applications of artificial intelligence (AI) in advancing the research field related to molecular mechanisms across different human populations or disease subgroups is indeed a crucial aspect to consider. AI, with its advanced algorithms for pattern recognition and data mining, can efficiently integrate and analyze massive multi-omics datasets (genomics, transcriptomics, proteomics, metabolomics, etc.). It can identify complex relationships and correlations among different omics layers, revealing novel molecular pathways and biomarkers that would be difficult to discern manually. The interpretation of AI-generated results requires expert knowledge to validate biological significance and avoid false discoveries. Data quality and standardization across different platforms and studies pose challenges for AI integration. AI-driven predictive models can leverage large-scale patient data to forecast disease progression, enabling early interventions. For treatment response, AI can identify patient subgroups that are most likely to benefit from specific therapies, personalizing treatment strategies. Model validation requires prospective studies with diverse patient populations to ensure accuracy and generalizability. Ethical considerations arise regarding the use of sensitive patient data for AI training and predictions. AI can rapidly screen vast chemical libraries to identify compounds that modulate the disulfidptosis pathway, accelerating drug discovery. It can also predict drug-target interactions and adverse effects, reducing the need for extensive preclinical testing. In vitro and in vivo validation of AI-predicted compounds is essential but time-consuming and costly. Drug specificity and selectivity remain challenges, especially for complex pathways like disulfidptosis. AI can analyze individual patients' molecular profiles to tailor treatment plans that maximize efficacy and minimize side effects. It can continuously monitor treatment response and adjust therapies accordingly, enabling precision medicine. Access to high-quality molecular data and advanced diagnostic tools may be limited, particularly in resource-constrained settings. The cost of personalized medicine can be prohibitive for some patients. AI can simulate experiments in silico, predicting outcomes and optimizing protocols before actual experiments, reducing time and resources. It can also identify experiments with the highest potential impact, minimizing the need for animal testing. In silico models are approximations and may not fully capture the complexity of biological systems. Validation of AI-optimized protocols in real-world experiments is necessary. AI-powered simulations can visualize and analyze cellular processes related to disulfidptosis at unprecedented resolution, providing insights into mechanisms and potential drug targets. They can explore hypothetical scenarios and test the effects of perturbations, facilitating hypothesis generation. The accuracy of simulations depends on the quality and completeness of the underlying biological models. Computational resources and expertise are required to develop and run complex simulations. It is crucial to emphasize that AI should not replace traditional experimental methods but rather complement them. A collaborative approach, where AI augments human expertise and experimentation, will maximize the benefits and mitigate the limitations. Researchers, clinicians, and data scientists must work together to ensure that AI applications are grounded in biological reality, ethical principles, and practical considerations.

Methodology:

While your animal studies provide crucial insights, consider discussing how some aspects of the research might be complemented or partially replaced by in silico methods in the future.

Thank you very much for your careful review and valuable feedback on my research work. While animal studies have been invaluable in providing crucial insights into various biological processes and diseases, there is a growing recognition that in silico methods, which rely on computational modeling and simulation, have the potential to complement or partially replace certain aspects of animal research in the future. One of the key advantages of in silico methods is their ability to rapidly test and evaluate hypotheses at a lower cost and with reduced ethical concerns compared to animal studies. This makes them particularly useful for preliminary screening of potential drug candidates or exploring the underlying mechanisms of complex diseases. Furthermore, in silico models can be tailored to specific biological systems or conditions, allowing researchers to investigate aspects of biology that may be difficult or impossible to study in animals. For example, computational models can be used to simulate the behavior of cells, tissues, or entire organisms under a wide range of conditions, providing insights into the complex interactions that occur within biological systems. However, it is important to note that in silico methods are not a panacea and have their own limitations. For example, they may not be able to fully capture the complexity and variability of biological systems, and they may require extensive validation using experimental data from animal or human studies. In the future, a combination of in silico methods and animal studies is likely to be the most effective approach for advancing biomedical research. By leveraging the strengths of both approaches, researchers can gain a more comprehensive understanding of biological processes and diseases, ultimately leading to the development of more effective treatments and therapies.

Broader impact:

Expand on how your findings on disulfidptosis might apply to other neurodegenerative diseases or cancer types beyond those directly studied.

Thank you very much for your careful review and valuable feedback on my research work. Research on "disulfidptosis" (disulfide-induced cell death), while potentially limited in its direct application to specific neurodegenerative diseases or cancer types, offers significant insights and promising avenues for the study of other disease categories. In Alzheimer's disease (AD), notable oxidative stress and mitochondrial dysfunction are prevalent, and disulfides, through their antioxidant or mitochondrial regulatory functions, may impact neuronal death in AD. One of the hallmarks of AD is the aggregation of β-amyloid and tau proteins, which may contribute to cell death. Investigating whether disulfides can influence the aggregation or clearance of these proteins could lead to novel therapeutic approaches for AD. PD is characterized by the progressive death of dopaminergic neurons in the substantia nigra. The mechanism of disulfide-induced cell death may share similarities with neuronal death in PD, rendering its study in PD models potentially valuable. Additionally, the aggregation of α-synuclein is another crucial pathological process in PD. Exploring whether disulfides affect the aggregation, propagation, or clearance of α-synuclein could provide new clues for PD treatment strategies. Huntington's disease (HD) arises from the expansion of CAG repeats in the huntingtin gene, leading to the aggregation of mutant huntingtin protein and neuronal death. Research into whether disulfides impact the toxicity of mutant huntingtin protein could reveal novel therapeutic targets.

Many types of cancer cells exhibit metabolic reprogramming, such as enhanced glycolysis and glutamine metabolism. Disulfides may influence cancer cell growth and survival by modulating these metabolic pathways. Based on the mechanism of disulfide-induced cell death, developing anticancer drugs targeting specific metabolic pathways could emerge as a new therapeutic strategy. Some cancer cells are resistant to apoptotic signals, enabling them to evade immune clearance and proliferate continuously. Investigating whether disulfides induce cancer cell death through non-apoptotic pathways, such as disulfide-induced cell death, could offer new methods to overcome apoptosis resistance. Cancer-associated fibroblasts (CAFs) play a crucial role in the tumor microenvironment, supporting cancer cell growth and invasion. Studying whether disulfides affect CAF function or their interactions with cancer cells could identify new therapeutic targets for cancer treatment. While research on the mechanism of disulfide-induced cell death in specific disease models is still preliminary, its potential antioxidant, mitochondrial regulatory, protein aggregation-modulating, and metabolic reprogramming properties offer broad prospects for the study of other neurodegenerative diseases and cancer types. Future research should further explore the mechanisms of disulfides in these diseases and assess their potential as therapeutic strategies.

Thank you very much for your careful review and valuable feedback on my research work. Further improve the quality of the manuscript.

Author Response File: Author Response.docx

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

Disulfidptosis has been implicated in various pathological conditions, including neurodegenerative disorders and cancer. Increasing evidence indicate an inverse association between overall cancer and most cancer types in patients with Parkinson's disease (PD). To elucidate the roles of disulfidptosis-related genes (DRGs) in PD and tumors and to assist personalized treatment for PD and cancer, this study investigated the function and expression of DRGs in PD and performed prognostic analysis of DRGs in PD using PD data samples obtained from GEO database, as well as analyzed the differential expression, prognosis, survival curves, immune infiltration, and single-cell types of DRGs in cancer using all cancer sample data obtained from UCSC database. The results showed that a total of 4 differential expression DRGs (DEDRGs) in PD were obtained, including ACTB, ACTN4, INF2 and MYL6. In the MPTP-induced PD mouse model, the expression of ACTB was decreased, while the expression of ACTN4, INF2 and MYL6 was increased. The expressions of these 4 DRGs also are closely related to pan-cancer and highly infiltrated in various immune cells in tumor tissues. There are some concerns for the present manuscript as listed in the following:

**L55: [7]-> L66: [12]: no Ref. 8,9,10,11

L130: In Vivo Experiment Verification-> In Vivo experiment verification

*L140: MPTP: give the full name and the source.

*L146: separated on ?% SDS PAGE

*L168: Using R software (version 3.6.4) calculated the expression differences between normal and tumor samples in each tumor using non-paired Wilcoxon Rank Sum and Signed Rank Tests for differential significance analysis -> editing

*L178: a previously published study in Cell ?[21]

**L304: Figure 3G: something wrong in the labels of the axis (i.e. ACTB ACTB ACTN4)

L308: α Braak α-synu-

L362: 2 tumor type -> 2 tumor types

**L651: References: inconsistent writing format for

(1) title: R20, 24, 40 (=43), 47, 52, 53,

(2) journal name: R6, 7, 9, 10, 11, 13, 14, 15, 16, 18, 19, 24, 25 -34,36, 40, 42, 45, 52, 58

**L742: Ref. 43= Ref. 40

Comments on the Quality of English Language

Minor editing of English language required.

Author Response

**L55: [7]-> L66: [12]: no Ref. 8,9,10,11

Thank you very much for your careful review and valuable feedback on my research work. Reference numbers have been added to the manuscript.

L130: In Vivo Experiment Verification-> In Vivo experiment verification

Thank you very much for your careful review and valuable feedback on my research work. It has been corrected in the manuscript.

*L140: MPTP: give the full name and the source.

Thank you very much for your careful review and valuable feedback on my research work. 1-methyl-4-phenyl-1,2,3,6-te-trahydropyridine

*L146: separated on ?% SDS PAGE

Thank you very much for your careful review and valuable feedback on my research work. 8%, 10%, and 12% SDS PAGE.

Thank you very much for your careful review and valuable feedback on my research work. Using R software (version 3.6.4), calculated the expression differences between normal and tumor samples in each tumor, using non-paired Wilcoxon Rank Sum and Signed Rank Tests for differential significance analysis.

*L178: a previously published study in Cell ?[21]

Thank you very much for your careful review and valuable feedback on my research work. The reference cited is [20].

**L304: Figure 3G: something wrong in the labels of the axis (i.e. ACTB ACTB ACTN4)

Thank you very much for your careful review and valuable feedback on my research work. In the analysis results, different gene ID correspond to the same gene, which has been annotated in the caption.

L308: α Braak α-synu-

Thank you very much for your careful review and valuable feedback on my research work. α has been deleted.

L362: 2 tumor type -> 2 tumor types

Thank you very much for your careful review and valuable feedback on my research work. It has been corrected in the manuscript.

**L651: References: inconsistent writing format for

(1) title: R20, 24, 40 (=43), 47, 52, 53,

(2) journal name: R6, 7, 9, 10, 11, 13, 14, 15, 16, 18, 19, 24, 25 -34,36, 40, 42, 45, 52, 58

Thank you very much for your careful review and valuable feedback on my research work. The reference format has been changed as required.

**L742: Ref. 43= Ref. 40

Thank you very much for your careful review and valuable feedback on my research work. Reference 43 has been deleted.

Author Response File: Author Response.docx

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

In this study, the authors tried to explore the role of disulfidptosis in the relationship between neurodegenerative diseases and tumorigenesis. The authors concluded that ACTB, ACTN4, INF2, and MYL6 were closely related to Parkinson disease and pan-cancer, which can be used as candidate genes for the diagnosis, prognosis and therapeutic biomarkers of neurodegenerative diseases and cancers.\

Comments:

The reviewer has some concerns as follows:

1. This study is interesting and novel. However, in this manuscript, there are still deficiencies in the organization, presentation, and interpretation of such a huge amount of data.

2. In the Introduction section, the reference citation should be clearly indicated. In Lines 44-45, 56-57, 63-64, and 69-71, the reference citations are lacking.

3. In the Methods section, (1) Animal subsection, the description for “male C57BL/6J mice aged 6–8 weeks, weighing between 25 and 30 grams” needs to be checked and revised. In general, the range of body weight for male C57BL/6J mice aged 6–8 weeks is 20-25 grams. (2) the sources for C57BL/6J mice and MPTP material need also be described. (3) in line 143, what is the black matter? What areas of brain tissues are used?

4. In the Results section, (1) in Figure 1A and B, the fonts are really too small to read clearly. This will lose the meaning of data presentation for this figure. (2) In the legend of Figure 2, the description of “The darker the blue color, the higher the expression level in the brain regions.” is shown. What does this sentence mean? (This is not a complete sentence) (3) in Figure 3G, what do the stages (stage 1-6) mean? How to decide these stages? Where is the stage 0 (line 308)? (4) in Figure 4A, the data seem to be not consistent. What is the sample size? (5) in Figure 4C, what is the cell types shown? Are the brain or cancer cells? (6) in lines 432-436, what does “the highest expression of MYL6 in various cells” mean? There are many cells shown. (7) In Figure 9A-D, the fonts are really too small to read clearly. (8) in Figure 11, the data for controls or non-tumor regions are lacking. (9) In Figure 12, the fonts are really too small to read clearly.

5. Why select the drugs of Cyclophosphamide and Ethinyl estradiol for testing as the therapeutic drugs and molecular docking targets? It needs to be explained.

6. In the Discussion section, in lines 625-628, the reference citation is lacking.

7. Overall, this manuscript needs to be majorly revised for writing, reference citation, methodology, and data presentation.

Comments on the Quality of English Language

An editing of English language is recommended.

Author Response

Comments:

The reviewer has some concerns as follows:

This study is interesting and novel. However, in this manuscript, there are still deficiencies in the organization, presentation, and interpretation of such a huge amount of data.

Thank you very much for your careful review and valuable feedback on my research work. We have revised the manuscript content.

In the Introduction section, the reference citation should be clearly indicated. In Lines 44-45, 56-57, 63-64, and 69-71, the reference citations are lacking.

Thank you very much for your careful review and valuable feedback on my research work. Due to personal error, the reference number was missing and has been added to the manuscript.

In the Methods section, (1) Animal subsection, the description for “male C57BL/6J mice aged 6–8 weeks, weighing between 25 and 30 grams” needs to be checked and revised. In general, the range of body weight for male C57BL/6J mice aged 6–8 weeks is 20-25 grams. (2) the sources for C57BL/6J mice and MPTP material need also be described. (3) in line 143, what is the black matter? What areas of brain tissues are used?

Thank you very much for your careful review and valuable feedback on my research work. The revisions have been tracked in the manuscript.

In the Results section, (1) in Figure 1A and B, the fonts are really too small to read clearly. This will lose the meaning of data presentation for this figure. (2) In the legend of Figure 2, the description of “The darker the blue color, the higher the expression level in the brain regions.” is shown. What does this sentence mean? (This is not a complete sentence) (3) in Figure 3G, what do the stages (stage 1-6) mean? How to decide these stages? Where is the stage 0 (line 308)? (4) in Figure 4A, the data seem to be not consistent. What is the sample size? (5) in Figure 4C, what is the cell types shown? Are the brain or cancer cells? (6) in lines 432-436, what does “the highest expression of MYL6 in various cells” mean? There are many cells shown. (7) In Figure 9A-D, the fonts are really too small to read clearly. (8) in Figure 11, the data for controls or non-tumor regions are lacking. (9) In Figure 12, the fonts are really too small to read clearly.

Thank you very much for your careful review and valuable feedback on my research work. (1) Figure 1 has been re uploaded. (2) The darker the blue color, the higher the protein expression level in the brain region. (3) In 2003, Braak et al. classified the pathological changes of Parkinson's disease into six stages based on the different deposition sites of alpha synuclein, the main component of Lewy bodies, and the time and order of PD pathology. They proposed that the onset of PD develops in an upward manner from the medulla oblongata to the cortex. When analyzing the GEO dataset, stage 0 is used as a healthy control and compared with stages 1-2 to obtain differentially expressed genes, so there is no stage 0. (4) Each result is repeated three times in parallel, with a sample size of 3. (5) The immunofluorescence image in Figure 4C was downloaded from the HPA database (https://www.proteinatlas.org/), and in order to display the distribution of proteins in cells, it is described as a human cell according to the database. (6) Thank you very much for pointing out the issues raised by the reviewers. They have been corrected in the manuscript. (7) Figure 9 has been re uploaded. (8) The expression of proteins in normal tissues has been increased in the supplementary materials. (9) Figure 12 has been re uploaded.

Why select the drugs of Cyclophosphamide and Ethinyl estradiol for testing as the therapeutic drugs and molecular docking targets? It needs to be explained.

Thank you very much for your careful review and valuable feedback on my research work. The gene-drug interaction network was constructed to identify potential new targets using the DGldb database (https://dgidb.org/). By predicting that only the ACTB gene has potential therapeutic drugs, analysis was conducted on Cyclophosphamide and Ethinyl estradiol.

In the Discussion section, in lines 625-628, the reference citation is lacking.

Thank you very much for your careful review and valuable feedback on my research work. References have been added.

Overall, this manuscript needs to be majorly revised for writing, reference citation, methodology, and data presentation.

Thank you very much for your careful review and valuable feedback on my research work. Based on your suggestions, we have revised the manuscript and your suggestions have greatly improved its quality.

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report (New Reviewer)

Comments and Suggestions for Authors

The authors responded to my comments well. Thank you.

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

This revised manuscript has a great improvement and the reviewer has no further comments.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This is an interesting manuscript, with a robust bioinformatics analysis, and it is important to mention that it also has several opportunity areas to improve

The main one is the lack of verification or validation of disulfoptosis in the generated model. Although it is verifiable that when animals were exposed to MPTP a PD was induced, it is still necessary to prove that disulfoptosis occurs. It is important to note that the authors, used the “veneer” package in R 4.2.0, to select the genes related to disulfoptosis. But If they analyzed the presence of SLC7A11 or reported the levels of disulfide molecules, it would ensure the relationship in sylico and in vivo.

Therefore, until this relation is proven, considering ACTB, ACTN4, INF2, and MYL6 as biomarkers for diagnosis in two diseases where cytological and histological criteria have great weight, is very risky.

In Figure 3, image G, it is not explained why for ACTB there are 6 data series on the horizontal axis (x) and different data series for ACTN4, INF2, and MYL6, considering that all were obtained from the data sheet GSE49036.

It is important to be cautious about the ACTB result, because on page 8, line 299 it is considered a tend to decrease, however, considering the WB results in the gels and graph, the most prudent approach is to mention them as a non-significant decrease.

Considering the premise that PD and cancer could present antagonistic behaviors regarding the expression of the selected genes; when reporting the analysis of observable mutations of the 4 genes, this objective is lost in the descriptive abundance.

Regarding the format, it is important to pay attention to the meaning of acronyms (page 1, line 7, PD), the correct use of capital letters (page 1, line 8,

disulfidptosis), and to watch out for the correct name of the cells “Eisinophils”

It would help if you were cautious regarding the classification of diseases, since mentioning in the introduction that “Cancer and neurodegenerative diseases are among the most chronic physiological disorders.”, the “physiological” term for many authors is controversial.

Comments on the Quality of English Language

It is a well-written manuscript, however, additional revision taking care of minor writing errors would help.

Author Response

This is an interesting manuscript, with a robust bioinformatics analysis, and it is important to mention that it also has several opportunity areas to improve

Thank you for your thoughtful review and constructive feedback on our manuscript. We appreciate your recognition of the robust bioinformatics analysis we conducted. We acknowledge the concern regarding the lack of verification or validation of disulfoptosis in our generated model. To address this, we conducted additional experiments to directly measure the levels of disulfide molecules and the expression of SLC7A11 in our model. This could provide a more definitive link between our in silico findings and in vivo mechanisms of disulfoptosis. Furthermore, we understand the caution associated with presenting ACTB, ACTN4, INF2, and MYL6 as potential biomarkers for the diagnosis of the diseases described. We agree that more validation is necessary, and we will explore experimental validation of these biomarkers in future experiments to strengthen our claims. We thank you again for your valuable insights, which will undoubtedly enhance the quality of our work. We will work diligently to incorporate these amendments in our revised manuscript.

Thank you for the question raised by the reviewer. In GEO (Gene Expression Omnibus) data, it is common for multiple probes to correspond to one gene. This is usually due to the following reasons: Different variants or transcripts of a gene: A gene can produce multiple variants or transcripts, and each probe may target a different transcript, resulting in multiple probes corresponding to the same gene. Highly homologous genes: Genes in certain gene families may have similar sequences, leading to cross reactivity of probes to different signals from these genes. At this point, multiple probes may correspond to the same or similar genes. Technical design: During the probe design phase, researchers may design multiple probes for the same gene to enhance the sensitivity and accuracy of detection. Differences in expression levels: Different probes may have different sensitivities, so when detecting the same gene, multiple probes may be used for a more comprehensive evaluation of expression levels.

Thank you for the question raised by the reviewer. We will rephrase the result as “non-significant decrease” based on your suggestion. This modification will better align with the Western blot experimental results and data charts we observed.

Thank you for the question raised by the reviewer. The mutation abundance of genes has been added to the text.

Regarding the format, it is important to pay attention to the meaning of acronyms (page 1, line 7, PD), the correct use of capital letters (page 1, line 8,disulfidptosis), and to watch out for the correct name of the cells “Eisinophils”

Thank you for the question raised by the reviewer. It has been corrected in the text.

Thank you for the question raised by the reviewer. The manuscript has been revised as follows: Cancer is one of the most common causes of death; meanwhile, the incidence and prevalence of central nervous system diseases are also very high. There exist complex biological connections between cancer and neurodegenerative diseases.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

Dear authors,

in your MS you describe disulfidptosis as a new target for parkinsons disease nd cancer. You employ analysis of DEDRGs derived from different datasets (GSE49036, 20163, 20164 gse7621, gse99039, GSE22491, and TCGA pan-cancer.

This begs my first question, why did you combine these two rather different pathologies? Please give a rationale.

Lines 106 to 124 are very unclear. This has to be completely rewritten according to the scheme

1) Data acquisition

2) Determination of differentially exporessed genes (linear models or otherwise)

3) Correction of the p-value (Benjamini Hochberg or other)

4) Further analysis

In summary, you should give a short workflow what you have done.

Lines 116 to 124: how did you determine hub genes?

Line 153 to 162: The PANCAN database contains three formats with 10535 samples, the TOIL RSEM fpkm, the norm_count or the tpm. All three can be seen as "normalized expression values". Which one did you use? Given you transformed the data using an offset of 0.001, I think it is either fpkm or tpm, both of which are not usable for the detection of differentially expressed genes with state of the art algorithms like DESeq2 or LIMMA. In any case, all three are already normalized (log2(value + 0.001 resp. 1) so that an additional log2 normalization must not be done.

In line 160 to 163 you describe the analysis of differentially expressed genes. Which packages did you use? Did you apply any multi testing correction? The state of the art method to determine DEGs from TCGA is outlined for example here: https://bioconductor.org/packages/release/bioc/html/TCGAbiolinks.html. Usually raw counts are dwnloaded and fed into a suitable algorithm like DESeq2 or LIMMA.

Lines 177 to 186. This section reads like a Standard Operating Procedure. Please refrase it to a proper materials and methods sction

Lines 187 to 196. Assessment of the immune infiltration. Why did you not use more advanced tools like Cibersort?

Given the serious flaws in the description of the methodology, assessment of the results is impossible.

In summary, this paper needs extensive rewriting and clarification, especially with regard to materials and methods.To me it seems that there are severe methodological shortcomings.

I also advise you to consult a bioinformatician with regard to the omics data.

Comments on the Quality of English Language

Not applicable.

Author Response

Dear authors,

This begs my first question, why did you combine these two rather different pathologies? Please give a rationale.

Thank you for the question raised by the reviewer. Combining Parkinson's disease (PD) and cancer as areas of focus for disulfidptosis research may initially seem unconventional due to their distinct pathophysiological mechanisms. However, several rationales can support this interdisciplinary approach:

Common Molecular Pathways: Both Parkinson's disease and cancer involve dysregulation of similar molecular pathways, including oxidative stress, inflammation, and cellular apoptosis. By examining disulfidptosis, which may influence these pathways, we can uncover shared mechanisms that could inform therapeutic strategies for both conditions.

Role of Cell Death Mechanisms: Disulfidptosis, as a form of regulated cell death, could play a role in both neurodegenerative processes seen in PD and tumorigenesis in cancer. Understanding how disulfidptosis operates in these contexts may yield insights into its potential as a target for interventions that could alleviate neurodegeneration or inhibit cancer cell proliferation.

Data Derivation from Diverse Datasets: The use of various datasets (e.g., GSE49036, GSE20163, GSE20164, GSE7621, GSE99039, GSE22491, and TCGA pan-cancer) enables a comprehensive analysis across different biological contexts. This approach can help identify common transcriptional or genetic signatures associated with disulfidptosis, providing a robust foundation for further exploration of this target in both diseases.

Translational Research Potential: By identifying disulfidptosis as a shared therapeutic target, it enhances the translational potential of the research, encouraging the development of novel treatments that could be beneficial across both Parkinson’s disease and cancer populations.

Holistic Understanding of Health: Disease processes often overlap, and studying them in conjunction allows for a more holistic understanding of health and disease. This integrative approach may lead to the discovery of novel biomarkers or therapeutic targets that address multiple conditions simultaneously.

Lines 106 to 124 are very unclear. This has to be completely rewritten according to the scheme

1) Data acquisition

2) Determination of differentially exporessed genes (linear models or otherwise)

3) Correction of the p-value (Benjamini Hochberg or other)

4) Further analysis

In summary, you should give a short workflow what you have done.

Thank you for the question raised by the reviewer. The manuscript has been revised as follows: PD-related differential expression genes (DEGs) were obtained from the gene4PD database (http://www.genemed.tech/gene4pd/). In addition, literature has been collected on genes associated with DRGs [15]. The “venneuler” package in R 4.2.0 software, ggplot2[3.3.6], VennDiagram[1.7.3] was adopted to draw the intersection of DEGs and DRGs, i.e., differential expression disulfidptosis-related genes (DEDRGs). The KEGG pathway and GO enrichment of DEDRGs were analyzed utilizing the gene4PD database, and used Cytoscape 3.10.1 to draw a network diagram. And analyze the average expression levels of DEDRGs utilizing the gene4PD database in different brain regions at different developmental period using “BrainSpan” under “Gene expression”, and the average expression level of DEDRGs in different tissue using “GTEx” under “Gene expression”.

Receiver operating characteristic (ROC) curve is a graphical representation of the diagnostic value of a gene in predicting a certain condition or disease. It plots the true positive rate (sensitivity) against the false positive rate (1-specificity) at different threshold values. To screen the diagnostic genes, we visually displayed the ROC curve analysis was performed, and the AUCs were calculated using the pROC package in R 4.2.1 software, pROC[1.18.0], ggplot2[3.3.6] to determine the predicted values of the hub genes. We selected PD datasets from the GEO database, including GSE49036, GSE20163, GSE20164, GSE7621, GSE99039, and GSE22491. Diagnostic genes were selected from the set using the criterion of AUC > 0.500.

Lines 116 to 124: how did you determine hub genes?

Thank you for the question raised by the reviewer. The hub genes (DEDRGs) are obtained by taking the intersection of PD differentially expressed genes (DEGs) and disulfidptosis-related genes (DRGs).

Thank you for the question raised by the reviewer. The expression form of cancer samples is fpkm, and the log2 (x+0.001) transformation is used to transform the expression value. Using R software (version 4.2.1), ggplot2[3.3.6], stats[4.2.1], car[3.1-0], we calculated the expression differences between normal and tumor samples in each tumor using non-paired Wilcoxon Rank Sum and Signed Rank Tests for differential significance analysis.

Thank you for the question raised by the reviewer. Using R software (version 4.2.1), ggplot2[3.3.6], stats[4.2.1], car[3.1-0], we calculated the expression differences between normal and tumor samples in each tumor using non-paired Wilcoxon Rank Sum and Signed Rank Tests for differential significance analysis.

Lines 177 to 186. This section reads like a Standard Operating Procedure. Please refrase it to a proper materials and methods sction

Thank you for the question raised by the reviewer. RNA sequencing (RNAseq) data were obtained from The Cancer Genome Atlas (TCGA) database, accessible via the GDC portal (https://portal.gdc.cancer.gov). We focused on cancer projects that exhibited significant differences in prognosis. To ensure robust analysis, we extracted the data in Transcripts Per Million (TPM) format alongside relevant clinical information sourced from a Cell article [20]. Normal samples and those with incomplete clinical data were systematically excluded to maintain the integrity of our analysis. To investigate the relationship between gene expression and patient survival, we employed Cox proportional hazards regression analysis using the 'survival' package in R (version 4.2.1). Prior to analysis, we conducted tests for the proportional hazards assumption to validate the suitability of the data for this statistical method. Survival curves derived from the Cox models were visualized using the 'survminer' and 'ggplot2' packages (version 3.3.6) to illustrate the survival outcomes across different cancer types. This approach facilitated an intuitive representation of survival differences based on the prognostic analysis. To further corroborate our findings regarding the variability in survival curves across cancer types, we utilized the GEPIA database (http://gepia.cancer-pku.cn/) for additional verification. This validation step was critical to substantiate our results and ensure the robustness of the observed survival disparities.

Lines 187 to 196. Assessment of the immune infiltration. Why did you not use more advanced tools like Cibersort?

Thank you for the question raised by the reviewer. The choice to use the ssGSEA algorithm instead of more advanced tools like CIBERSORT can be attributed to several factors:

Data Type and Sample Size: ssGSEA can be more suitable for smaller sample sizes or datasets where the number of samples is limited. It computes pathway enrichment scores for each sample independently, which can be advantageous when dealing with datasets that lack sufficient samples for robust deconvolution, such as those often found in clinical studies.

Simplicity and Interpretability: ssGSEA provides a straightforward scoring system for immune infiltration, where for each immune cell type, an enrichment score is calculated based on the expression of specific marker genes. This simplicity can sometimes lead to more interpretable results, especially for those who may not have extensive experience with more complex deconvolution methods.

Flexibility: ssGSEA does not require a reference gene expression profile for cell types, making it more versatile. Users can define their own gene sets based on their specific research needs. This is particularly useful if the available reference profiles do not perfectly align with the biological context under investigation.

Robustness to Noise: The ssGSEA method is less sensitive to noise in the data compared to some other methods, making it potentially more reliable for analyses on datasets that may have varying levels of quality and technical noise.

Focus on Functional Analysis: If the primary goal is to perform functional enrichment analysis of immune cells rather than strictly deconvoluting cell types, ssGSEA’s scoring approach might be more aligned with the research objectives.

In summary, while tools like CIBERSORT offer more advanced deconvolution capabilities, ssGSEA provides certain advantages in terms of flexibility, interpretability, and applicability to specific types of data and research questions. The selection of tools ultimately hinges on the specific goals of the analysis, the characteristics of the dataset, and the expertise of the researchers involved.

Author Response File: Author Response.docx

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Dear authors,

Thank you for your speedy reply.

gene4pd: Gene4PD allows for data analysis or the download of differentially expressed genes as a text file. If you did indeed analyze the data, you must provide the parameters for this analysis based on a short workflow; in the case of using the DEG file, state that. However, one has to make sure the p-values are corrected for multi-testing. Otherwise, the number of false positives will go through the roof. The paper should mention the adjustment method to enable the reader to assess how strict the adjustment was.

Determination of hub genes: A hub gene is central to a network, not the intersection between DEGs and DRGs. This should be clearly written as "we determined relevant genes by extracting DRGs out of DEGs" or something similar, not called "hub gene."

PANCAN expression data:

If you used the fpkm values as a starting point for your analysis, this has to be stated in the MS since there are several different measures for gene expression (for example, fpkms, tpms, and counts). However, using fpkm and some statistical tests differs from how it's done. Counts are the starting point for a DEG analysis of NGS data in state-of-the-art workflows (such as LIMMA and DESeq2). These packages employ models specifically tailored for the analysis of count-based data. PANCAN also provides counts, and the TCGA allows a straightforward way to analyze these data via TCGABiolinks. FPKMs are only used if nothing better is available (https://support.bioconductor.org/p/9154302/).

Use more advanced deconvolution methods. Cybersort is not the only algorithm. For example, MCP-counter or xCell employ both GSEA-like approaches. quanTIseq is specifically designed for cancer samples. The advantage of using published deconvolution algorithms is that they have been validated to fit the purpose.

Flexibility: ssGSEA requires at least a set of marker genes.

Availability of reference gene sets: If a particular reference gene matrix is unavailable, one can create or modify an existing one from publicly available data, as all research groups developing such algorithms have done.

Finally, the argument for focusing on a more functional analysis is moot since one can always combine functional analysis methods with deconvolution methods.

With all due respect, I again strongly advise you to consult a bioinformatician with experience in analyzing such data.

Best regards,

Comments on the Quality of English Language

Not applicable

Author Response

Thank you for the question raised by the reviewer. We selected p-value<0.05 as the PD differentially expressed gene. Due to the limited number of disulfidptosis-related genes (DRGs), our inclusion criteria for PD DEGs were relatively lenient, which is a limitation of this study and has been mentioned at the end of the manuscript.

Thank you for the question raised by the reviewer. The intersection of DEGs and DRGs, i.e., differential expression disulfidptosis-related genes (DEDRGs).

PANCAN expression data:

Thank you for the question raised by the reviewer. In the pan cancer expression data, we analyzed the differential expression of DEDRGs in pan cancer using the Sanger Box database (http://sangerbox.com/).

Flexibility: ssGSEA requires at least a set of marker genes.

Finally, the argument for focusing on a more functional analysis is moot since one can always combine functional analysis methods with deconvolution methods.

With all due respect, I again strongly advise you to consult a bioinformatician with experience in analyzing such data.

Thank you for the question raised by the reviewer. Your points about exploring more advanced deconvolution methods are well taken. Indeed, methods like MCP-counter and xCell offer valuable alternatives to Cybersort and utilize gene set enrichment analysis (GSEA)-like approaches that can enrich our understanding of cellular composition in complex tissues. Similarly, quanTIseq's design specifically for cancer samples highlights the importance of using tools tailored to the specific biology of the system under study.

The flexibility of ssGSEA is a double-edged sword; while it allows researchers to customize and adapt the analysis, it does depend on the availability and selection of appropriate marker genes. As you mentioned, if a reference gene set is not readily available, researchers can generate or refine datasets from existing public resources, which many groups have successfully done. This adaptability is crucial in the ever-evolving landscape of genomics.

Your suggestion to integrate functional analysis with deconvolution methods is particularly insightful. Combining these approaches can provide a more comprehensive understanding of the biological context, leading to richer interpretations of the data.

Finally, your recommendation to seek guidance from a bioinformatician is prudent. Collaborating with someone experienced in this field can enhance the analysis and interpretation of results, ensuring that the methodologies employed align with best practices and current standards in the field. Thank you for your valuable input!

Round 3

Reviewer 2 Report

Comments and Suggestions for Authors

Dear authors,

just one point. Re PANCAN. In the MS you state that you downloaded the fpkm values,, transformed them using the log2 transformation, and then applied some R packages. Now you state that the sangerbox database was used. Aside that the website is in Chinese - which makes a proper assessment difficult - you should mention the usage of this tool. However, according to the paper, the underlying algorithm in this toolbox is LIMMA. LIMMA can handle fpkm values - when certain precautions are taken. In any case, if counts are available, these must be used (see for example here: https://www.biostars.org/p/389815/). Without a proper, methodologically sound analysis, the results are not reliable. With all respect, please analyze the data according to state-of-the art methods. As long as this is not done, I can not assess the validity of the results.

best,

Thomas

Comments on the Quality of English Language

Author Response

just one point. Re PANCAN. In the MS you state that you downloaded the fpkm values,, transformed them using the log2 transformation, and then applied some R packages. Now you state that the sangerbox database was used. Aside that the website is in Chinese - which makes a proper assessment difficult - you should mention the usage of this tool. However, according to the paper, the underlying algorithm in this toolbox is LIMMA. LIMMA can handle fpkm values - when certain precautions are taken. In any case, if counts are available, these must be used (see for example here: https://www.biostars.org/p/389815/). Without a proper, methodologically sound analysis, the results are not reliable. With all respect, please analyze the data according to state-of-the art methods. As long as this is not done, I can not assess the validity of the results.

Thank you very much for the critical questions raised by the reviewers in the article. We acknowledge that a series of analyses were conducted on pan cancer using the Sanger Box online analysis website, and the methods were organized according to the methods provided on the website. It is worth mentioning that although this website is in Chinese, it includes a pan cancer dataset and only requires manual input of the target gene to obtain results such as differential expression and prognosis of the target gene in pan cancer. The experimental methods are now reorganized. Thank you for your feedback, which has further improved the quality of our manuscript. We take ACTB as an example:

DEDRGs differential expression of Pan-cancer, The expression values of genes in pan cancer are shown in Supplementary Table S1.

DEDRGs prognostic analysis of Pan-cancer

Author Response File: Author Response.docx

Article Menu

Disulfidptosis: A New Target for Parkinson’s Disease and Cancer

Further Information

Guidelines

MDPI Initiatives

Follow MDPI