Functional characterisation of rare variants in genes encoding the MAPK/ERK signalling pathway identified in long-lived Leiden Longevity Study participants

Article activity feed

1. 

   Version published to 10.1101/2023.06.01.541708v2 on bioRxiv

   May 23, 2024
2. 

   Review Commons

   May 14, 2024

   Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Reply to the reviewers

   We would like to thank the reviewers for their insightful comments and suggestions, which helped us to streamline and improve our manuscript. Below you can find a detailed response to each of their raised points. In short, we redid most of our experiments to get cleaner data, added some additional experiments (based on suggestions of the reviewers) to strengthen our conclusions, and removed the fly-related data to make the manuscript more straightforward. Moreover, we have combined our Results and Discussion section to adhere to the formatting of EMBO Reports.

   Reviewer #1

   Major comments

   * The mESC data on the various mutations would be more convincing …

   Note: This response was posted by the corresponding author to Review Commons. The content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Reply to the reviewers

   We would like to thank the reviewers for their insightful comments and suggestions, which helped us to streamline and improve our manuscript. Below you can find a detailed response to each of their raised points. In short, we redid most of our experiments to get cleaner data, added some additional experiments (based on suggestions of the reviewers) to strengthen our conclusions, and removed the fly-related data to make the manuscript more straightforward. Moreover, we have combined our Results and Discussion section to adhere to the formatting of EMBO Reports.

   Reviewer #1

   Major comments

   * The mESC data on the various mutations would be more convincing if derived from two lines, respectively, as in the case of Phe1112Leu NF1 mutation.*

   We agree with the reviewer that it would have been more convincing if we would have a second mESC line harbouring the Asp633Tyr variant in RAF1. However, we were not successful in creating such an additional line. Moreover, it would not be feasible, both financially and time-wise, to redo all our experiments with this additional line. However, we have unpublished data that shows that the transgenic mESC line harbouring the Asp633Tyr variant in RAF1 shows clustering with and similar effects as several transgenic mESC lines harbouring other genetic variants in different genes from a connected pathway (which we plan to publish in another manuscript), making us less concerned that the observed effects are caused by random off-target effects.

   * The results concerning ERK1/2 phosphorylation in mESC are actually reflecting the basal MAPK/ERK activity of cells maintained in normal growth medium. It would be important to check the MAPK/ERK activation by specific stimuli like EGF upon starvation in the mESC lines harboring the rare variants.*

   We thank the reviewer for this suggestion. Based on this comment, we performed additional experiments in which we stimulated our transgenic mESC lines with both EGF and insulin. These experiments showed similar results as the ones we performed in normal growth medium (see updated Figure EV3), strengthening our conclusions that our variants indeed alter MAPK/ERK signalling pathway activity. Moreover, we could additionally show that they also down-regulate phosphorylation of p70 S6K (see updated Figure 2C), indicating reduced mTORC1 activity, which has previously been associated with increased lifespan in different model organisms.

   * According to the KEGG pathway analysis in Fig 4D PI3K-Akt signaling is activated in both NF1 and RAF1 variants. Because of the well-known cross-talk of PI3K/Akt with MAPK/ERK signaling it would strengthen the paper if PI3K-Akt signaling is analyzed, for example by determining the phosphorylation of Akt.*

   Based on comment 10 of reviewer 2, we re-analysed the proteomics data and treated each of the genetic variants separately. Although the PI3K-AKT signalling pathway does not show a significant enrichment in the separate groups, we did measure the phosphorylation of AKT and p70 S6K (see reply to comment 2) to probe the effects of the variants on insulin/IGF-1 signalling. We indeed found that both variants up-regulate phosphorylation of AKT at S473 in normal growth medium (although the effect of the NF1 variant is clearly stronger) and down-regulate phosphorylation of AKT at S308 after insulin stimulation, while we observed a RAF1-specific down-regulation of phosphorylation of AKT at S473 after insulin stimulation (see updated Figure 2B and 2D).

   * Some players in the MAPK/ERK signaling pathway are upregulated, some are down-regulated in mutant NF1 or mutant RAF1 cells, but it is not clear what the net effect of all these changes is on MAPK/ERK signaling. However, what ultimately matters are changes in down-stream gene expression. To really determine the effect of the mutations on MAPK/ERK signaling it would be necessary to perform more detailed transcriptome analysis and especially check the expression of longevity-controlling transcription factors, such as SKN-1, ETS and FOXO.*

   We thank the reviewer for this very helpful suggestion. We performed an additional experiment in which we looked at the effect of our variants on the transcription of mammalian orthologs of the lifespan-associated transcription factors that belong to the SKN-1, ETS and FOXO family. We specifically focussed on the subset of ETS transcription factors that have been linked with lifespan regulation in fruit flies, given the known relation with MAPK/ERK signalling. In line with our findings from the proteomics, we indeed found consistent (i.e. Nfe2l2, Foxo3, Etv1, and Etv6) as well as opposing effects (i.e. for Ets1, Ets2, and Etv4) on the expression levels of these transcription factors between our cell lines (see new Figure 4). Based on this we concluded that both cell lines show reduced MAPK/ERK signalling activity.

   * The authors are discussing a gain of function effect of the variants on the activity of RAF1 and of NF1, based on the ERK1/2 phosphorylation data from mESC. Since the variants are residing in protein domains important for the respective protein function (Tubulin-binding domain of NF1 and C-terminus of RAF1, important for its interaction with 14-3-3 proteins, respectively), the authors could speculate on how the mutations might affect the respective protein activity. Furthermore, the data could be strengthened by directly testing the activity of RAF1 or Ras.*

   We have now added some text in which we speculate on the potential effects of our genetic variants (i.e. gain- or loss-of-function). Since we were mostly interested in the (shared) downstream effects of the variants, we decided to focus on this instead of the activity of Raf1 (for which good assays are also lacking).

   * Since there is no consistent effect of the investigated mutations and their effects on MAPK/ERK signaling in mESC and no consistent effect on life-span in flies, wouldn´t one have to conclude that the pipeline for functional characterization is actually not working? Along that line, if introducing putative human life-extending mutations in RAF and NF1 in flies leads to lethality in one case and a shortened life-span in the other, doesn´t that proof the model is not suitable to draw conclusions about human mutations in flies?*

   We have decided to remove the fly data from our manuscript to make the message more straightforward. We also realised that the lifespan-associated effects of the protein in flies had been contributed by its role in the adenylate cyclase-cAMP-protein kinase A pathway and not MAPK/ERK signalling (PMID: 17369827). Hence, we were not sure if the reduced lifespan effects we observed could be attributed to the role of Nf1 on MAPK/ERK signalling, especially since we did not observe any effects on phosphorylation of ERK1/2 in the flies.

   Minor issues

   * Introduction, last paragraph.*

   The sentence "Notably.... is very long and could be changed to two sentences.

   We have adapted this.

   * Results, paragraph "Generation of mESCs..."*

   It only becomes clear in the discussion that the AN3-12 cells get diploid after a while and that the human donors were heterozygous. This should be mentioned already here.

   We have adapted this.

   * Results, paragraph "Generation and characterization of transgenic flies.."*

   What is the wDah background?

   As mentioned above, we have removed all the fly data from our manuscript.

   * The dimer consisting of RTK and a GPCR in the simplified illustration of the MAPK/ERK signaling pathway in Fig.1(B) is misleading, it is probably supposes to be a RTK?*

   We have adapted this.

   Reviewer #2

   Major comments

   * The NF1 variant and RAF1 variant have different outcomes regarding ERK phosphorylation. Then, how can long-lived family members share these variants?*

   This is indeed a good point. However, now that we redid most of our experiments, we are able to show that most of the effects of the variants are consistent, especially when looking at the main effects on MAPK/ERK signalling. However, the proteomics and transcriptomic analyses still show some opposing and diverging effects. Hence, we speculate that this likely indicates that there are multiple ways in which genetic variants could influence cellular processes/phenotypes associated with healthy ageing and there is not a single molecular mechanism explaining it all.

   * The two variants in mESCs showed contradictory results on MAPK/ERK pathway. In addition, fruit fly didn't recapitulate the results of mESC experiments. How can the authors conclude these variants are causally linked to longevity?*

   See our reply to comment 1 and to major comment 6 of reviewer 1.

   * Figure 2C, The authors should correct the statistical test (they used a T-test for 4 sample data set).*

   We have adapted this.

   * Figure 2C, Is NF1 and MEK1/2 expression altered? What about pMEK1/2 expression? The mechanistic link between NF1 mutations and ERK phosphorylation is speculative.*

   We thank the reviewer for this suggestion. We have now also added data on the phosphorylation of MEK1/2, which showed consistent results with that of ERK1/2.

   * Figure 2C, The loading looks very variable. The authors should use fluorescently labelled antibodies for multiplexing. This way, the phospho signal and total protein can be quantified on a single blot.*

   We have redone all our western blots and now normalised to calnexin, since we realised that vinculin was relatively unstable and therefore not the best reference protein to use. All data looks consistent now.

   * Figure 2D, Loading control, Vinculin, is variable. Based on vinculin expression, total ERK expression was increased in RAF1 Asp633Tyr variant. It could affect the amount of pERK. The authors should show whether the authors loaded the equal amount of proteins using stain-free as in Fig. 5D.*

   See reply to comment 5.

   * Figure 2D, What about total RAF1, MEK1/2 expression and pMEK1/2? I was wondering whether phosphorylation of ERK is increased via RAF1-MEK pathway. The link between RAF1 mutation and ERK phosphorylation is mechanistically speculative.*

   See reply to comment 4. The expression of Raf1 itself is provided in Figure 3A (i.e. the proteomic dataset) and is differentially influenced by both variants.

   * Figure 3C and D, It doesn't look like dramatic improvement, especially since the curves run in parallel. The authors should corroborate the findings using an assay that is independent on the cellular metabolism, e.g. cell survival or proliferation using Incucyte*

   We redid our experiments using the Incucyte® Live-Cell Analysis System and focussed on a stressor (i.e. hydroxyurea) that showed consistent effects across experiments (see updated Figure 5B). The stressors we used previously did not work so well in this system. We also measured proliferation in normal growth medium (see updated Figure 5A). These results indicate that the proliferation of the NF1Phe1112Leu variant mESC lines was increased, while that of the RAF1Asp633Tyr variant mESC line was decreased under normal growth conditions. Moreover, the RAF1Asp633Tyr variant showed improved resistance to replication stress, while this was not the case for the NF1Phe1112Leu variant.

   * Figure 4, To figure out global phosphorylation changes induced by the variants, I suggest the authors perform phospho-proteomics*

   We agree with the reviewer that it would be very nice to perform phospho-proteomics. However, this is still relatively expensive and our proteomics facility mentioned that such measurements are likely not yet robust and sensitive enough to get reliable estimates of specific phosphorylation sites.

   * Figure 4C and D, NF1 variants and RAF1 variants have an opposite effect on phosphorylation of ERK. Why did the authors investigate the shared upregulated or downregulated proteins between two variants? How can they share TFs with the MAPK/ERK signaling pathway?*

   See our reply to comment 1 and to major comment 4 of reviewer 1.

   * Figure 4, The conclusion of this figure 4 is not clear to me.*

   We have now updated the text in the Results and Discussion section to make this clearer.

   Minor comments

   * The ultimate goal of aging research will be healthy aging. In LLS, were all long-lived individuals healthy? Do the authors have additional clinical parameters?*

   There is only limited clinical data available for the long-lived individuals from the Leiden Longevity Study (PMID: 27374409), so we were not able to thoroughly asses this (also because there is no appropriate control group to compare them to). For the sequencing, we focussed on the individuals that had the longest survival within their families, but we cannot rule out that some of them were relatively unhealthy.

   * Overall, most western blot figures do not look like being representative of the quantification results. The authors need better representative western blot figures*

   We have repeated all our western blot experiments and updated our figures to show the most representative images.

   * Figure 3, is there a difference in cell proliferation ability/viability between WT and the NF1/RAF1 variants?*

   See reply to major comment 8.

   * Figure 4A, How many replicates were used here?*

   We used 4 technical replicates per cell line. We have now added this information to the Figure legend and the text in the Methods section.

   * Figure 4A, The authors should provide the rationale for the cutoff they used: fold change and p-value/FDR?*

   We have now added this information to the Figure legend and the text in the Results and Discussion section.

   * Figure 5C and D, Did mutant flies die due to aging or due to any disease?*

   As mentioned before, we have removed all the fly data from our manuscript.

   Read the original source
3. 

   Review Commons

   May 14, 2024

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #2

   Evidence, reproducibility and clarity

   Summary

   As part of a meta-study using the whole genome sequencing data from Leiden Longevity Study (LLS), the author identified uncommon genetic variants in MAPK/ERK signaling pathway which are potentially associated with human longevity. To characterize these gene variants, the authors employed CRISPR/Cas9 genome-edited mouse embryonic stem cells (mESCs) and fruit flies. Paradoxically, the variants in NF1 and RAF1 (both associated with increased longevity) have functionally opposite effect on activity of MAPK/ERK pathway in vitro. Nf1 variant in flies has no effect on MAPK/ERK pathway, however, it leads to deleterious consequences …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #2

   Evidence, reproducibility and clarity

   Summary

   As part of a meta-study using the whole genome sequencing data from Leiden Longevity Study (LLS), the author identified uncommon genetic variants in MAPK/ERK signaling pathway which are potentially associated with human longevity. To characterize these gene variants, the authors employed CRISPR/Cas9 genome-edited mouse embryonic stem cells (mESCs) and fruit flies. Paradoxically, the variants in NF1 and RAF1 (both associated with increased longevity) have functionally opposite effect on activity of MAPK/ERK pathway in vitro. Nf1 variant in flies has no effect on MAPK/ERK pathway, however, it leads to deleterious consequences such as shorter lifespan, delayed developmental time, and decreased locomotor activity. Due to the contradictory results of the in vitro experiments and the in vivo fly model, it is difficult to conclude that the rare genetic variants the identified, are linked to longevity.

   Major comments

   1. The NF1 variant and RAF1 variant have different outcomes regarding ERK phosphorylation. Then, how can long-lived family members share these variants?
   2. The two variants in mESCs showed contradictory results on MAPK/ERK pathway. In addition, fruit fly didn't recapitulate the results of mESC experiments. How can the authors conclude these variants are causally linked to longevity?
   3. Figure 2C, The authors should correct the statistical test (they used a T-test for 4 sample data set).
   4. Figure 2C, Is NF1 and MEK1/2 expression altered? What about pMEK1/2 expression? The mechanistic link between NF1 mutations and ERK phosphorylation is speculative.
   5. Figure 2C, The loading looks very variable. The authors should use fluorescently labelled antibodies for multiplexing. This way, the phospho signal and total protein can be quantified on a single blot.
   6. Figure 2D, Loading control, Vinculin, is variable. Based on vinculin expression, total ERK expression was increased in RAF1 Asp633Tyr variant. It could affect the amount of pERK. The authors should show whether the authors loaded the equal amount of proteins using stain-free as in Fig. 5D.
   7. Figure 2D, What about total RAF1, MEK1/2 expression and pMEK1/2? I was wondering whether phosphorylation of ERK is increased via RAF1-MEK pathway. The link between RAF1 mutation and ERK phosphorylation is mechanistically speculative.
   8. Figure 3C and D, It doesn't look like dramatic improvement, especially since the curves run in parallel. The authors should corroborate the findings using an assay that is independent on the cellular metabolism, e.g. cell survival or proliferation using Incucyte
   9. Figure 4, To figure out global phosphorylation changes induced by the variants, I suggest the authors perform phospho-proteomics
   10. Figure 4C and D, NF1 variants and RAF1 variants have an opposite effect on phosphorylation of ERK. Why did the authors investigate the shared upregulated or downregulated proteins between two variants? How can they share TFs with the MAPK/ERK signaling pathway?
   11. Figure 4, The conclusion of this figure 4 is not clear to me.

   Minor comments

   1. The ultimate goal of aging research will be healthy aging. In LLS, were all long-lived individuals healthy? Do the authors have additional clinical parameters?
   2. Overall, most western blot figures do not look like being representative of the quantification results. The authors need better representative western blot figures
   3. Figure 3, is there a difference in cell proliferation ability/viability between WT and the NF1/RAF1 variants?
   4. Figure 4A, How many replicates were used here?
   5. Figure 4A, The authors should provide the rationale for the cutoff they used: fold change and p-value/FDR?
   6. Figure 5C and D, Did mutant flies die due to aging or due to any disease?

   Significance

   Filtering out meaningful rare variants in MAPK/ERK pathway from long-lived individuals is an interesting and promising approach. However, the link between the NF1/RAF1 variants and longevity is still unclear. The authors were not able to explain the contradictory results from the NF1 and RAF1 mutant mESCs. In addition, the fly model did not support the in vitro mESC results. The authors need to provide more mechanistic insight into the impact of the variants on MAPK signaling. This part of the study is very superficial. Overall, the story seems a bit premature.

   Advance: The authors identify rare mutations affecting the ERK pathway in long-lived family members.

   Audience: Basic researchers who are interested in signaling and aging.

   Read the original source
4. 

   Review Commons

   May 14, 2024

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   Hinterding et al. present a manuscript where they characterize rare variants in genes found in long-lived families. The authors concentrated on the MAPK/ERK signaling pathway, because they argued that this pathway has an established role in life-span determination. The rare variants were introduced in mouse embryonic stem cells and in fruit flies and their effects on the MAPK/ERK pathway and on life-span was studied. The authors conclude they established a pipeline for the functional characterization and potential validation of rare genetic variants.

   The topic is very interesting and the approach is original and novel. However, the …

   Note: This preprint has been reviewed by subject experts for Review Commons. Content has not been altered except for formatting.

   Learn more at Review Commons

   ---

   Referee #1

   Evidence, reproducibility and clarity

   Hinterding et al. present a manuscript where they characterize rare variants in genes found in long-lived families. The authors concentrated on the MAPK/ERK signaling pathway, because they argued that this pathway has an established role in life-span determination. The rare variants were introduced in mouse embryonic stem cells and in fruit flies and their effects on the MAPK/ERK pathway and on life-span was studied. The authors conclude they established a pipeline for the functional characterization and potential validation of rare genetic variants.

   The topic is very interesting and the approach is original and novel. However, the results are in part preliminary and contradictory and the conclusions are overstated. Before publication, we suggest to address a number of issues.

   1. The mESC data on the various mutations would be more convincing if derived from two lines, respectively, as in the case of Phe1112Leu NF1 mutation.
   2. The results concerning ERK1/2 phosphorylation in mESC are actually reflecting the basal MAPK/ERK activity of cells maintained in normal growth medium. It would be important to check the MAPK/ERK activation by specific stimuli like EGF upon starvation in the mESC lines harboring the rare variants.
   3. According to the KEGG pathway analysis in Fig 4D PI3K-Akt signaling is activated in both NF1 and RAF1 variants. Because of the well-known cross-talk of PI3K/Akt with MAPK/ERK signaling it would strengthen the paper if PI3K-Akt signaling is analyzed, for example by determining the phosphorylation of Akt.
   4. Some players in the MAPK/ERK signaling pathway are upregulated, some are down-regulated in mutant NF1 or mutant RAF1 cells, but it is not clear what the net effect of all these changes is on MAPK/ERK signaling. However, what ultimately matters are changes in down-stream gene expression. To really determine the effect of the mutations on MAPK/ERK signaling it would be necessary to perform more detailed transcriptome analysis and especially check the expression of longevity-controlling transcription factors, such as SKN-1, ETS and FOXO.
   5. The authors are discussing a gain of function effect of the variants on the activity of RAF1 and of NF1, based on the ERK1/2 phosphorylation data from mESC. Since the variants are residing in protein domains important for the respective protein function (Tubulin-binding domain of NF1 and C-terminus of RAF1, important for its interaction with 14-3-3 proteins, respectively), the authors could speculate on how the mutations might affect the respective protein activity. Furthermore, the data could be strengthened by directly testing the activity of RAF1 or Ras.
   6. Since there is no consistent effect of the investigated mutations and their effects on MAPK/ERK signaling in mESC and no consistent effect on life-span in flies, wouldn´t one have to conclude that the pipeline for functional characterization is actually not working? Along that line, if introducing putative human life-extending mutations in RAF and NF1 in flies leads to lethality in one case and a shortened life-span in the other, doesn´t that proof the model is not suitable to draw conclusions about human mutations in flies?

   Minor issues

   1. Introduction, last paragraph. The sentence "Notably.... is very long and could be changed to two sentences.
   2. Results, paragraph "Generation of mESCs..." It only becomes clear in the discussion that the AN3-12 cells get diploid after a while and that the human donors were heterozygous. This should be mentioned already here.
   3. Results, paragraph "Generation and characterization of transgenic flies.." What is the wDah background?
   4. The dimer consisting of RTK and a GPCR in the simplified illustration of the MAPK/ERK signaling pathway in Fig.1(B) is misleading, it is probably supposes to be a RTK?

   Referees cross-commenting

   Reviewer #2 provides a fair and balanced review and seems to have pretty much the same concerns as I do, namely that there are too many inconsistencies in the experiments to conclude that the identified candidate genes are longevity genes.

   Significance

   The concept of the study, to look for gene variants in long-lived families, is novel and highly interesting. It should be relevant for a broad audience interested in aging, longevity and the underlying mechanisms.

   Strengths:

   • identification of potentially long-life associated gene variants in humans

   Limitations:

   • final outcome on MAPK/ERK signaling not analyzed (downstream genes)
   • investigated gene variants don´t show consistent pattern
   • use of flies as model for the analysis of human longevity gene variants not convincing since one mutation is lethal, the other life shortening

   My expertise is cell biology, aging, senescence. I co-reviewed the paper with my postdoc who worked on MAPK/ERK signaling for many years.

   Read the original source
5. 

   Version published to 10.1101/2023.06.01.541708v1 on bioRxiv

   Jun 1, 2023