Investigating PRDM8 DNA Methylation in Peripheral Tissues in Borderline Personality Disorder: Association with Symptom Severity but Not Adverse Childhood Experiences

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Excellent work, well written, well documented, and well executed. I congratulate the authors on a remarkable achievement. The analysis is clear and concise, with a well-defined question that has been answered in a simple yet meaningful way.

I cannot assess the biological analysis, as it lies outside my field of expertise; however, both the psychopathological assessment and the statistical analyses are exemplary. I particularly appreciate the inclusion of normality tests prior to conducting further analyses—something I now rarely see.

The results and discussion are also superb. Without any unnecessary verbosity, the discussion elegantly assigns to future studies the task of clarifying what this work could not address. In my view, this has been accomplished perfectly.

Author Response

Comment 1:

Excellent work, well written, well documented, and well executed. I congratulate the authors on a remarkable achievement. The analysis is clear and concise, with a well-defined question that has been answered in a simple yet meaningful way.

I cannot assess the biological analysis, as it lies outside my field of expertise; however, both the psychopathological assessment and the statistical analyses are exemplary. I particularly appreciate the inclusion of normality tests prior to conducting further analyses—something I now rarely see.

The results and discussion are also superb. Without any unnecessary verbosity, the discussion elegantly assigns to future studies the task of clarifying what this work could not address. In my view, this has been accomplished perfectly.

Answer: We would like to thank the reviewer for this very appreciative report and the positive comments about our work.

Reviewer 2 Report

Comments and Suggestions for Authors

This case control study compared DNAm patterns of PRDM8 in patients with BPD versus health controls and after exposure to treatment while considering their histories of ACE. The study results suggested that PRDM8 DNAm was associated with BPD and treatment outcomes but not with ACE. The research findings are interesting and raised more questions to be addressed. Correctly, the authors unscored the complexities in studying genetics, trauma and psychopathology. The study was well described but the authors should have addressed the following: 
1.    Did study participants with BPD also have co-morbid PTSD? Patients with BPD comorbid with PTSD are known to demonstrate more symptom severity and PTSD may mediate the relationship between borderline psychopathology and PRDM8 DNAm. 

2.    Overall, the authors should have discussed the issue of co-morbid disorders and how future research should account for co-morbid disorders.

Author Response

1.    Did study participants with BPD also have co-morbid PTSD? Patients with BPD comorbid with PTSD are known to demonstrate more symptom severity and PTSD may mediate the relationship between borderline psychopathology and PRDM8 DNAm.

Unfortunately, information about comorbidities such as PTSD is not available for the investigated cohort. We agree, that this is a limitation of our study as comorbidities could influence the observed findings. We now included this in our limitation section (page 10, line 416) as follows:

Furthermore, data on prenatal trauma as well as comorbidities such as PTSD or major depressive disorder were not available, which made disentanglement of pre- and postnatal adversity impossible and may have mediated the association between BPD and DNAm in PRDM8.

2.    Overall, the authors should have discussed the issue of co-morbid disorders and how future research should account for co-morbid disorders.

We agree with the reviewer that the issue of co-morbid disorders should have been addressed. We now include this in our discussion (page 9, line 339) as follows:

BPD frequently co-occurs with other psychiatric conditions—such as posttraumatic stress disorder (PTSD), major depressive disorder, anxiety disorders, substance abuse, and eating disorders. BPD patients suffering from comorbid diseases frequently display a more pronounced symptom severity and a worse prognosis. In addition, they show poorer response to treatment [5]. However, these comorbidities may not only influence clinical measures, but also confound or modulate the epigenetic signatures associated with BPD, including DNA methylation. Unfortunately, information about comorbidities were not available for our sample. However, future studies should include this information to gain more insight about the complex nature of BPD.

The following sentence has been included into the outlook (page 10, line 431):

Furthermore, future studies should stratify subjects according to comorbid disorders such as PTSD or major depressive disorder in order to refine our understanding of how comorbidity modulates the epigenetic landscape in BPD and support more personalized therapeutic approaches. 

Reviewer 3 Report

Comments and Suggestions for Authors

BRAIN SCIENCE- 3829432

1. The research paper titled “Investigating PRDM8 DNA Methylation in Peripheral Tissues in Borderline Personality Disorder: Association with Symptom Severity but not Adverse Childhood Experiences” investigates the epigenetic regulation, specifically DNA methylation patterns of the gene PRDM8, in individuals with Borderline Personality Disorder (BPD). The study seeks to examine differences in PRDM8 DNA methylation between BPD patients and healthy controls, assess changes following therapy (Dialectical Behavioral Therapy), and explore the association of methylation levels with exposure to childhood adversity (Adverse Childhood Experiences, ACE). Additionally, the research aims to determine whether PRDM8 methylation could serve as a biomarker for disease state and therapeutic response in BPD.
2. The study is significant since it addresses a timely and relevant question: how gene-environmental interactions, particularly epigenetic mechanisms, contribute to psychiatric disorders like BPD.
3. The manuscript is well crafted and neatly organized.
4. The author should provide gene and protein expression data of PRDM8
5. The author should justify that blood methylation patterns can be indicative, explain how these reflect central nervous system changes an important point given the psychiatric focus. Because some epigenetic modifications are tissue specific here it is brain tissue. Are there any studies discussed PRDM8 expression or DNA methylation pattern in brain cells or tissue if so, discuss it.
6. The findings are correlational, it remains unclear whether methylation changes are causes, consequences, or epiphenomena linked to BPD pathology or therapy effects.
7. The lack of prenatal trauma data limits understanding of pre- versus postnatal influences on methylation.
8. The author should justify the comment:  Without knowing the gender composition, it is difficult to ascertain whether the study findings can be applied broadly across both males and females. Gender differences in epigenetic patterns, trauma exposure, and clinical presentation may influence the outcomes, particularly in studies investigating DNA methylation and psychiatric disorders.
9. The lack of replication in saliva may reflect tissue specificity but also raises questions about generalizability and biological relevance.
10. The author should discuss the functional impact of PRDM8 methylation changes on gene expression which could provide mechanical insights.
11. The author should provide pictorial representation of the overall findings.

I recommend this article for acceptance after the author addresses all the above points.

Comments for author File: Comments.pdf

Author Response

1. The research paper titled “Investigating PRDM8 DNA Methylation in Peripheral Tissues in Borderline Personality Disorder: Association with Symptom Severity but not Adverse Childhood Experiences” investigates the epigenetic regulation, specifically DNA methylation patterns of the gene PRDM8, in individuals with Borderline Personality Disorder (BPD). The study seeks to examine differences in PRDM8 DNA methylation between BPD patients and healthy controls, assess changes following therapy (Dialectical Behavioral Therapy), and explore the association of methylation levels with exposure to childhood adversity (Adverse Childhood Experiences, ACE). Additionally, the research aims to determine whether PRDM8 methylation could serve as a biomarker for disease state and therapeutic response in BPD.
2. The study is significant since it addresses a timely and relevant question: how gene-environmental interactions, particularly epigenetic mechanisms, contribute to psychiatric disorders like BPD.
3. The manuscript is well crafted and neatly organized.

We would like to thank the reviewer for the positive comments regarding our manuscript.

1. The author should provide gene and protein expression data of PRDM8

We agree with the reviewer that it is a limitation of our study that gene and protein expression data are not available. Only EDTA whole blood samples and saliva samples were collected and both are not suitable to analyze gene and protein expression. Therefore, they were only used to extract genomic DNA. As participants have been recruited and samples have been collected between 2013 and 2016, we can also not make up for the omission today. However, future studies should definitely include gene and protein expression data. We already mentioned the missing gene expression data in our limitations section, but now also included protein expression. Furthermore, we already stated that future studies should include gene expression data. But we now expanded that to protein expression in our outlook section.

The sentence in the limitations section (page 10, line 418) now reads.

Lastly, gene and protein expression levels were not assessed, which hinders interpretation of possible functional consequences of the observed DNAm patterns.

And in the outlook section (page 10, line 424) it now reads:

Building on our insights, while also considering the limitations of our study, future research should be encouraged to explore whether differential PRDM8 DNAm corresponds to gene and protein expression changes or involves other regulatory mechanisms, such as H3K9 methylation at BPD-related loci.

1. The author should justify that blood methylation patterns can be indicative, explain how these reflect central nervous system changes an important point given the psychiatric focus. Because some epigenetic modifications are tissue specific here it is brain tissue. Are there any studies discussed PRDM8 expression or DNA methylation pattern in brain cells or tissue if so, discuss it.

As DNA methylation patterns are tissue- and cell-type specific, we cannot assume that the changes we observe in whole blood directly reflect DNAm alterations in the brain. However, in psychiatric epigenetics we are faced with the problem, that the tissue of interest – the brain – is not available for molecular analysis in living individuals. We can therefore not conclude any epigenetic correlations between BPD and PRDM8 DNAm in the brain based on our results.  Furthermore, there are no previous reports of PRDM8 DNAm in the brain available, however, several studies reported gene expression of PRDM8 in brain and neuronal tissue. We included this in the discussion (page 8, line 314):

Based on our results, we cannot conclude any epigenetic correlations between BPD and PRDM8 DNAm in the brain. However, in psychiatric epigenetics we are faced with the problem, that the tissue of interest – the brain – is not available for molecular analysis in living individuals. We can therefore only report associations in peripheral tissue without providing information about mechanisms in the brain. Furthermore, there are no reports of PRDM8 DNAm in brain or neuronal tissue available, making it impossible to speculate whether the observed differences in whole blood PRDM8 DNAm between BPD patients and healthy control individuals could be reflective of the situation in the brain. However, previous reports indicate that PRDM8 is expressed in neuronal and brain tissue throughout development and in adulthood, indication a potential role of PRDM8 in the nervous system [30, 31, 33-35].

1. The findings are correlational, it remains unclear whether methylation changes are causes, consequences, or epiphenomena linked to BPD pathology or therapy effects.

Thank you for this valuable remark. We included this in our discussion (page 10, line 396) as follows:

Furthermore, we have to acknowledge that our findings are purely correlational and we cannot establish any causal relationship between PRDM8 DNAm and BPD or treatment outcome. As we did not perform a longitudinal study, we can also not decipher whether PRDM8 DNAm might be a consequence of the disorder or whether PRDM8 DNAm is merely an epiphenomenon coincidental to the pathology or treatment of BPD.

1. The lack of prenatal trauma data limits understanding of pre- versus postnatal influences on methylation.

We acknowledge this limitation and included it in our discussion (page 10, 392), which now reads:

Thus, while our data do not suggest a direct association of ACE and PRDM8 DNAm, the broader relationship between pre- and postnatal trauma, PRDM8 epigenetic regulation, and BPD psychopathology remains to be further elucidated, especially as the lack of prenatal trauma data in our sample hinders the decoding of effects of pre- vs. postnatal trauma on PRDM8 DNAm.

1. The author should justify the comment:  Without knowing the gender composition, it is difficult to ascertain whether the study findings can be applied broadly across both males and females. Gender differences in epigenetic patterns, trauma exposure, and clinical presentation may influence the outcomes, particularly in studies investigating DNA methylation and psychiatric disorders.

We would like to point out that our sample included male and females and was balanced for sex, however our male sample was very small (7 male BPD patients and 7 healthy control individuals). We can therefore not draw any conclusion about potential gender differences. We agree that this should be addressed in future studies to investigate the effect of sex on DNAm, but also trauma exposure and symptom severity and presentation in more detail. We included this in the limitation section (page 10, line 409) as follows:

Third, the overall sample size was relatively small, especially the male subsample, which may have obscured effects after controlling for multiple testing. In general, although we included males and females in our study, our study design while balanced for sex and age between healthy control individuals and BPD patients, was not suitable to decipher sex-specific effects. As DNAm as well as clinical symptom presentation is sex-specific, gender might play an important role and should be investigated further.

In the outlook (page 10, 427) we now stated that:

Upcoming studies would profit from larger, well-balanced cohorts including male and females, that enable a stratified analysis by both pre- and postnatal trauma exposure.

1. The lack of replication in saliva may reflect tissue specificity but also raises questions about generalizability and biological relevance.

As epigenetic patterns are usually tissue specific, we hypothesize that this is the reason for discrepancies between whole blood and saliva. However, this also indicates that the results are not generalizable across tissues. We discussed this as answered above (remark 5). We would not agree that this implies a lack of biological relevance.

1. The author should discuss the functional impact of PRDM8 methylation changes on gene expression which could provide mechanical insights.

As we unfortunately did not analyze gene or protein expression, we can not make any assumptions about the functional impact of PRDM8 DNAm. However, we included possible functional consequences in the discussion (page 9, line 351):

Its function as a methyltransferase makes PRDM8 a putative modifier of the histone code, possibly regulating downstream processes relevant for BPD. Specifically, PRDM8 has been shown to reduce transcription of target genes via H3K9 methylation [30, 63]. Although intragenic DNA hypomethylation is generally associated with a decrease in gene expression, intragenic PRDM8 hypermethylation (specifically, cg27242132 and cg19409579) has been reported to correspond to lower PRDM8 gene expression [64]. Accordingly, a theoretical upregulation of PRDM8 gene expression in BPD would potentially correspond to an increase of H3K9 methylation levels. Of note, increased H3K9 trimethylation has been associated with early life stress in animal models [65], and in depressed individuals committing suicide [66]. However, PRDM8 expression changes have not been consistently linked to its DNAm status [67]. Generally, intragenic DNAm is less extensively characterized than promoter DNAm and may influence not only transcription itself but also alternative splicing regulation [68-70]. Further, gene and protein expression and the interplay of complementary epigenetic regulatory mechanisms such as H3K9 trimethylation were not assessed in this study. Future research is warranted to investigate the relationship of PRDM8 DNAm and its expression, as well as to elucidate potential PRDM8-related molecular mechanisms underlying BPD.

1. The author should provide pictorial representation of the overall findings.

As we already included DNAm comparisons in whole blood and saliva between healthy control individuals, BPD patients before and after therapy as well as displayed the relationship between DNAm and symptom severity before therapy in both tissues, we are of the opinion that the main findings are already represented by graphs. As associations between ACE and DNAm were not detectable, and there was no significant association between CTQ scores and DNAm, we decided not to show this in an additional figure.

I recommend this article for acceptance after the author addresses all the above points.

We would like to thank the reviewer for this positive statement and hope we addressed al concerns adequately so that the manuscript will now be recommended for funding.