Statin-mediated reduction in mitochondrial cholesterol primes an anti-inflammatory response in macrophages by upregulating Jmjd3
Curation statements for this article:-
Curated by eLife
eLife assessment
The manuscript by Salloum and colleagues shows that cholesterol-lowering statins can reduce mitochondrial cholesterol and impact epigenetic programs in macrophages. The findings could be valuable for understanding statin-mediated anti-inflammatory functions in macrophages. The major claims describing new mechanisms by which statins may regulate macrophage function via epigenetic programming are partially supported by the data presented.
This article has been Reviewed by the following groups
Listed in
- Evaluated articles (eLife)
- Biochemistry and Chemical Biology (eLife)
Abstract
Stains are known to be anti-inflammatory, but the mechanism remains poorly understood. Here we show that macrophages, either treated with statin in vitro or from statin-treated mice, have reduced cholesterol levels and higher expression of Jmjd3 , a H3K27me3 demethylase. We provide evidence that lowering cholesterol levels in macrophages suppresses the ATP synthase in the inner mitochondrial membrane (IMM) and changes the proton gradient in the mitochondria. This activates NFkB and Jmjd3 expression to remove the repressive marker H3K27me3. Accordingly, the epigenome is altered by the cholesterol reduction. When subsequently challenged by the inflammatory stimulus LPS (M1), both macrophages treated with statins in vitro or isolated from statin-treated mice in vivo , express lower levels pro-inflammatory cytokines than controls, while augmenting anti-inflammatory Il10 expression. On the other hand, when macrophages are alternatively activated by IL4 (M2), statins promote the expression of Arg1 , Ym1 , and Mrc1 . The enhanced expression is correlated with the statin-induced removal of H3K27me3 from these genes prior to activation. In addition, Jmjd3 and its demethylase activity are necessary for cholesterol to modulate both M1 and M2 activation. We conclude that upregulation of Jmjd3 is a key event for the anti-inflammatory function of statins on macrophages.
Article activity feed
-
-
-
Author Response
Reviewer #1 (Public Review):
The manuscript by Salloum and colleagues examines the role of statin-mediated regulation of mitochondrial cholesterol as a determinant of epigenetic programming via JMJD3 in macrophages.
Key strengths of the work include:
- Mechanistic analysis of how statin treatments can remodel the mitochondrial membrane content via cholesterol depletion which in turn affects JMJD3 levels is a novel concept.
- Use of RNA-seq and ATAC-seq data provides an avenue for unbiased analysis of the statin effects.
- Use of methyl-cyclodextrin (MCD) alongside statins increases the robustness of the findings and the use of NFKB inhibitors suggests a mechanistic role for NFKB.
The conclusions are only partially supported by the presented data:
- There is a lack of any in vivo studies that are required to …
Author Response
Reviewer #1 (Public Review):
The manuscript by Salloum and colleagues examines the role of statin-mediated regulation of mitochondrial cholesterol as a determinant of epigenetic programming via JMJD3 in macrophages.
Key strengths of the work include:
- Mechanistic analysis of how statin treatments can remodel the mitochondrial membrane content via cholesterol depletion which in turn affects JMJD3 levels is a novel concept.
- Use of RNA-seq and ATAC-seq data provides an avenue for unbiased analysis of the statin effects.
- Use of methyl-cyclodextrin (MCD) alongside statins increases the robustness of the findings and the use of NFKB inhibitors suggests a mechanistic role for NFKB.
The conclusions are only partially supported by the presented data:
- There is a lack of any in vivo studies that are required to demonstrate that the concentrations of statins used to induce epigenetic programming of macrophages are physiologically relevant. There have been numerous studies that have examined the anti-inflammatory effects of statins but there is significant debate and controversy regarding the in vivo relevance. Much of the in vivo effects of statins are achieved via changes in systemic cholesterol levels but the direct effects on macrophages are not clear.
More discussion on this issue has been added (P9, line 9-33)
- "Statins" is used globally and it is unclear which statins were used, which doses of statins, and the treatment durations.
Names of the statins have been added for the individual experiments in the figure legends.
- The RNA-seq, ATAC-seq, and selected H3K27 ChIP only show a snapshot of the results without leveraging the power of unbiased analysis. Such an unbiased analysis could show whether the examined genes are indeed the most relevant targets of statins.
(a). Data are now analyzed with unsupervised GSEA, i.e. on all differentially expressed genes, both up and down, to identify the most significantly altered pathways. TNFa signalling via NF-aB came out on top (Fig. 1 A), similar to our conclusion from previous analyses.
- CCCP depletion can have broad toxic effects and it is difficult to interpret specific roles of ATP synthase from potentially toxic mitochondrial uncoupling.
CCCP within the dosages used in this study has no detectable toxicity. An MTT test was performed and added (Supplementary Fig. 5).
Reviewer #3 (Public Review):
The manuscript by Salloum et al., titled "Statin-mediated reduction in mitochondrial cholesterol primes an anti-inflammatory response in macrophages by upregulating JMJD3" reports an extensive characterization of the mechanisms underlying the anti-inflammatory role of statins using different in vitro studies. Based on these approaches, the authors observed that cholesterol reduction in response to statin treatment alters mitochondrial function and they identify JMJD3 as a potential critical driver of macrophage anti-inflammatory phenotype. Overall, the study is interesting and provides new findings that could shed light on the molecular effects of statins in these cells, but a number of issues remain confusing, and the experimental design is, on some occasions, not rigorous enough to support the drawn conclusions.
Major issues:
- Focus on JMJD3 is justified by the authors as it was among the 40 genes commonly up-regulated in macrophages exposed to statin or methyl--cyclodextrin (MCD) by RNA-Seq analysis. However, this analysis has not been presented in the manuscript and it is unclear what genes (apart from JMJD3) might play an important role in the response of these cells. A detailed characterization of both up- and down-regulated genes in these experimental conditions and a better justification for JMJD3 are required to fully support further analysis.
a. RNA-seq data from statin- and MCD-treated macrophages was re-analyzed by unsupervised Gene Set Enrichment Analysis (GSEA) (Fig. 1 A & B), which includes all differentially expressed genes, up and down, by cholesterol reduction. The conclusion is identical to the previous analysis, i.e. NF-kB is the top pathway activated by cholesterol reduction. The analysis in last version, which used a different program, is now moved to Supplementary Fig. 1.
b. ATAC-seq data was similarly re-analyzed with GSEA (Fig. 6 A). Again, NF-kB is the top pathway activated by cholesterol reduction (Fig. 6 A, b). Examples of the lineups between ATAC-Seq peaks and RNA-seq peaks have been added (Fig. 6 B).
c. RNA-seq data from LPS-stimulated macrophages with or without statins is also re-analyzed. Gene Ontology (GO) analysis of genes showing decreased expression upon statin treatment revealed that statins primarily suppress inflammatory processes (Fig. 7 A, b), while genes involved in cellular homeostatic functions were upregulated (Fig. 7 A, c).
- In the same line, Figures 6A and B fail to fully describe the changes found by ATAC-seq and RNA-seq. A more comprehensive analysis of these three datasets (together with previous RNA-seq studies) would help to obtain a better understanding of overlapping dysregulated genes (not only those found up-regulated) and what other epigenetic modifying factors might be involved.
See response to reviewer #1, 3. Also response to reviewer #2, 3.
- In Figure 6C and Supplementary Figure 7, it would be noteworthy to also measure the gene expression of Kdm6a/UTX homolog Kdm6c/UTY, as it has been shown to lack demethylate H3K27me3 demethylase activity due to mutations in the catalytic site of the Jumomji-C-domain.
Kdm6c/UTY in human is a male specific histone demethylase (PMID: 24798337). As statins are not known for sex-biases, this demethylase is not likely to play a role here.
- The use of rather unspecific treatments such as MG-132 (proteasome inhibitor) and GSKj4 (inhibitor of both JMJD3 and UTX) may distort the results observed and might elude their correct interpretation. To avoid this limitation, additional silencing and/or overexpression experiments are currently needed.
Jmjd3 knockdown experiments have been added to complement the glutamine-free and GDKj4 experiments (Fig. 8, C).
- Figure 3 and Supplementary Figure 3 seem to be duplicated, please correct them. Moreover, for a better representation of these data, please include representative Seahorse profile figures of each experimental condition in these figures.
Sorry for the error. It is corrected (Fig. 3, BMDMs).
- As stated by the authors, macrophage phenotype is much more complex than M1/M2 polarization. In this view, assessing a very limited set of genes (i.e, Il-1, IL-10, TNF, IL-6, IL-12, Arg1, Ym1, Mrc1) appears to be inappropriate. A meaningful number of markers must be added.
Yes, this is complex, and it would good if we could assess more genes for this purpose. M1/M2 polarization is relatively poorly defined, in terms of genes expressed. We used a list of genes that most tested in literature. For example, Nat Immunol. 2017 Sep;18(9):985-994.
- For accurate quantification of H3K27me3 global levels, please add immunoblotting against histone H3 in Supplementary Figure 1. Will look for it. H3 and H327me3 could not do in the same plots. It would involve stripping, which we do not trust.
No-stripping was the exact reason we didn’t use H3 as loading control. Comparison between separate plots could be another source of error. In addition, we would like to control for the effective cholesterol reduction in these cells by p-Creb. Whole cell lysates were used for western blotting, with actin as control for cell numbers.
-
eLife assessment
The manuscript by Salloum and colleagues shows that cholesterol-lowering statins can reduce mitochondrial cholesterol and impact epigenetic programs in macrophages. The findings could be valuable for understanding statin-mediated anti-inflammatory functions in macrophages. The major claims describing new mechanisms by which statins may regulate macrophage function via epigenetic programming are partially supported by the data presented.
-
Reviewer #1 (Public Review):
The manuscript by Salloum and colleagues examines the role of statin-mediated regulation of mitochondrial cholesterol as a determinant of epigenetic programming via JMJD3 in macrophages.
Key strengths of the work include:
1. Mechanistic analysis of how statin treatments can remodel the mitochondrial membrane content via cholesterol depletion which in turn affects JMJD3 levels is a novel concept.
2. Use of RNA-seq and ATAC-seq data provides an avenue for unbiased analysis of the statin effects.
3. Use of methyl-cyclodextrin (MCD) alongside statins increases the robustness of the findings and the use of NFKB inhibitors suggests a mechanistic role for NFKB.
The conclusions are only partially supported by the presented data:
1. There is a lack of any in vivo studies that are required to demonstrate that the …
Reviewer #1 (Public Review):
The manuscript by Salloum and colleagues examines the role of statin-mediated regulation of mitochondrial cholesterol as a determinant of epigenetic programming via JMJD3 in macrophages.
Key strengths of the work include:
1. Mechanistic analysis of how statin treatments can remodel the mitochondrial membrane content via cholesterol depletion which in turn affects JMJD3 levels is a novel concept.
2. Use of RNA-seq and ATAC-seq data provides an avenue for unbiased analysis of the statin effects.
3. Use of methyl-cyclodextrin (MCD) alongside statins increases the robustness of the findings and the use of NFKB inhibitors suggests a mechanistic role for NFKB.
The conclusions are only partially supported by the presented data:
1. There is a lack of any in vivo studies that are required to demonstrate that the concentrations of statins used to induce epigenetic programming of macrophages are physiologically relevant. There have been numerous studies that have examined the anti-inflammatory effects of statins but there is significant debate and controversy regarding the in vivo relevance. Much of the in vivo effects of statins are achieved via changes in systemic cholesterol levels but the direct effects on macrophages are not clear.
2. "Statins" is used globally and it is unclear which statins were used, which doses of statins, and the treatment durations
3. The RNA-seq, ATAC-seq, and selected H3K27 ChIP only show a snapshot of the results without leveraging the power of unbiased analysis. Such an unbiased analysis could show whether the examined genes are indeed the most relevant targets of statins.
4. CCCP depletion can have broad toxic effects and it is difficult to interpret specific roles of ATP synthase from potentially toxic mitochondrial uncoupling.
-
Reviewer #2 (Public Review):
In this study, the authors pursue a line of inquiry related to the impacts of cholesterol depletion on macrophage gene expression. The authors find that depletion of cholesterol with either statins or methyl-cyclodextrin induces robust gene expression changes, including changes to JMJD3 expression, an epigenetic regulator.
The authors then seek to dissect the mechanistic determinants of the regulation of JMJD3 in macrophages converging on a metabolic regulation hypothesis that requires mitochondrial activity.
A strength of the paper is the use of multiple macrophage cell models and multiple tools for perturbation to improve the rigor of their conclusions. A weakness of the paper is that it relies heavily on chemical approaches without ever using genetic tools to confirm that their conclusions can be …
Reviewer #2 (Public Review):
In this study, the authors pursue a line of inquiry related to the impacts of cholesterol depletion on macrophage gene expression. The authors find that depletion of cholesterol with either statins or methyl-cyclodextrin induces robust gene expression changes, including changes to JMJD3 expression, an epigenetic regulator.
The authors then seek to dissect the mechanistic determinants of the regulation of JMJD3 in macrophages converging on a metabolic regulation hypothesis that requires mitochondrial activity.
A strength of the paper is the use of multiple macrophage cell models and multiple tools for perturbation to improve the rigor of their conclusions. A weakness of the paper is that it relies heavily on chemical approaches without ever using genetic tools to confirm that their conclusions can be supported using an alternative approach, and when perturbing metabolic pathways as described here, it is difficult to understand how the entire cell state has changed. In fact, the unique focus on JMJD3 without utilizing a control set of genes to show that the impacts of these metabolic perturbations are specific to JMJD3 makes it hard to understand if this is a truly specific pathway for JMJD3 or a general cellular health change.
The authors make an interesting claim in the early part of their manuscript about the potential for statins to regulate the epigenome which they show; however, in the present presentation, it is unclear if this is related to the JMJD3 effect or a separate form of regulation.
This work has the potential to contribute to an improved understanding of the impact of statins on immune function.
-
Reviewer #3 (Public Review):
The manuscript by Salloum et al., titled "Statin-mediated reduction in mitochondrial cholesterol primes an anti-inflammatory response in macrophages by upregulating JMJD3" reports an extensive characterization of the mechanisms underlying the anti-inflammatory role of statins using different in vitro studies. Based on these approaches, the authors observed that cholesterol reduction in response to statin treatment alters mitochondrial function and they identify JMJD3 as a potential critical driver of macrophage anti-inflammatory phenotype. Overall, the study is interesting and provides new findings that could shed light on the molecular effects of statins in these cells, but a number of issues remain confusing, and the experimental design is, on some occasions, not rigorous enough to support the drawn …
Reviewer #3 (Public Review):
The manuscript by Salloum et al., titled "Statin-mediated reduction in mitochondrial cholesterol primes an anti-inflammatory response in macrophages by upregulating JMJD3" reports an extensive characterization of the mechanisms underlying the anti-inflammatory role of statins using different in vitro studies. Based on these approaches, the authors observed that cholesterol reduction in response to statin treatment alters mitochondrial function and they identify JMJD3 as a potential critical driver of macrophage anti-inflammatory phenotype. Overall, the study is interesting and provides new findings that could shed light on the molecular effects of statins in these cells, but a number of issues remain confusing, and the experimental design is, on some occasions, not rigorous enough to support the drawn conclusions.
Major issues:
1. Focus on JMJD3 is justified by the authors as it was among the 40 genes commonly up-regulated in macrophages exposed to statin or methyl--cyclodextrin (MCD) by RNA-Seq analysis. However, this analysis has not been presented in the manuscript and it is unclear what genes (apart from JMJD3) might play an important role in the response of these cells. A detailed characterization of both up- and down-regulated genes in these experimental conditions and a better justification for JMJD3 are required to fully support further analysis.
2. In the same line, Figures 6A and B fail to fully describe the changes found by ATAC-seq and RNA-seq. A more comprehensive analysis of these three datasets (together with previous RNA-seq studies) would help to obtain a better understanding of overlapping dysregulated genes (not only those found up-regulated) and what other epigenetic modifying factors might be involved.
3. In Figure 6C and Supplementary Figure 7, it would be noteworthy to also measure the gene expression of Kdm6a/UTX homolog Kdm6c/UTY, as it has been shown to lack demethylate H3K27me3 demethylase activity due to mutations in the catalytic site of the Jumomji-C-domain.
4. The use of rather unspecific treatments such as MG-132 (proteasome inhibitor) and GSKj4 (inhibitor of both JMJD3 and UTX) may distort the results observed and might elude their correct interpretation. To avoid this limitation, additional silencing and/or overexpression experiments are currently needed.
5. Figure 3 and Supplementary Figure 3 seem to be duplicated, please correct them. Moreover, for a better representation of these data, please include representative Seahorse profile figures of each experimental condition in these figures.
6. As stated by the authors, macrophage phenotype is much more complex than M1/M2 polarization. In this view, assessing a very limited set of genes (i.e, Il-1, IL-10, TNF, IL-6, IL-12, Arg1, Ym1, Mrc1) appears to be inappropriate. A meaningful number of markers must be added.
7. For accurate quantification of H3K27me3 global levels, please add immunoblotting against histone H3 in Supplementary Figure 1. -