Interplay between PML NBs and HIRA for H3.3 dynamics following type I interferon stimulus

  1. Constance Kleijwegt
  2. Florent Bressac
  3. Coline Seurre
  4. Wilhelm Bouchereau
  5. Camille Cohen
  6. Pascale Texier
  7. Thomas Simonet
  8. Laurent Schaeffer
  9. Patrick Lomonte
  10. Armelle Corpet

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    In this manuscript, of interest to the fields of animal immunity and epigenetics, the authors investigate the crosstalk between PML Nuclear Bodies and HIRA, a member of the H3.3 histone chaperone complex, during inflammatory stress. This study raises interesting perspectives on how availability of HIRA could be regulated by PML Nuclear Bodies for histone deposition onto interferon-stimulated genes, which in turn, could be relevant for immune-response mediated gene regulation.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Promyelocytic leukemia Nuclear Bodies (PML NBs) are nuclear membrane-less organelles physically associated with chromatin underscoring their crucial role in genome function. The H3.3 histone chaperone complex HIRA accumulates in PML NBs upon senescence, viral infection or IFN-I treatment in primary cells. Yet, the molecular mechanisms of this partitioning and its function in regulating histone dynamics have remained elusive. By using specific approaches, we identify intermolecular SUMO-SIM interactions as an essential mechanism for HIRA recruitment in PML NBs. Hence, we describe a role of PML NBs as nuclear depot centers to regulate HIRA distribution in the nucleus, dependent both on SP100 and DAXX/H3.3 levels. Upon IFN-I stimulation, PML is required for interferon-stimulated genes (ISGs) transcription and PML NBs become juxtaposed to ISGs loci at late time points of IFN-I treatment. HIRA and PML are necessary for the prolonged H3.3 deposition at the transcriptional end sites of ISGs, well beyond the peak of transcription. Though, HIRA accumulation in PML NBs is dispensable for H3.3 deposition on ISGs. We thus uncover a dual function for PML/PML NBs, as buffering centers modulating the nuclear distribution of HIRA, and as chromosomal hubs regulating ISGs transcription and thus HIRA-mediated H3.3 deposition at ISGs upon inflammatory response.

Article activity feed

  1. Version published to 10.7554/elife.80156 on eLife
  2. eLife

    Author Response

    Reviewer #1 (Public Review):

    The manuscript "Interplay between PML NBs and HIRA for H3.3 dynamics following type I interferon stimulus" by Kleijwegt and colleagues describes a study that's set out to explore the details of the PML-HIRA axis in H3.3 deposition at ISGs upon IFN-I stimulation. First, the authors establish that HIRA colocalized at PML NBs upon TNFa and TNFb treatment. This process is SUMO-dependent and facilitated by at least one of the identified SIM domains of HIRA. Next, the authors set out to determine whether interferon responsive genes (ISGs) are dependent on HIRA or PML. By knocking-down either HIRA or PML, only an effect on ISGs was observed when PML was knocked down. In fact, immune-FISH showed that PML NBs are in close proximity of ISGs upon TNFb treatment. To address the histone chaperone function of HIRA, the deposition of the replication-independent H3.3 on ISGs is tested. In specific, the enrichment of H3.3 across the ISG gene body. ChIP-seq data (Fig 5B) showed an enrichment around the TES, whereas qPCR (Fig 5A) showed less convincing enrichment (for details see below). When either HIRA or PML are knocked down, a mild loss of H3.3 enrichment was observed (Fig 5E). Interestingly, when HIRA is sequestered away from PML NBs by Sp100, an increased enrichment of H3.3 was observed. To understand the interplay between H3.3 deposition and HIRA's role in this process in the presence of PML NBs, H3.3 was overexpressed. Two population of cells were observed: low or high levels of H3.3. In the former, HIRA formed foci and the latter, HIRA did not form foci. Surprisingly, when HIRA is overexpressed, PML NBs form in the absence of TNFb. Finally, a two-sided model is proposed, where PML NBs is required for ISG transcription promoting H3.3 loading. The second side is that PML NBs function as a "storage center" for HIRA to regulate its availability.

    Overall, it the model is intriguing, but the data presented seems insufficient to support the current claims.

    We thank the reviewer for his/her constructive comments. We want to point out that there is a confusion in the reviewer's statement (highlighted in red here above) between TNFb and IFNb, because it is IFNb that was mostly used in our study. We suppose it is a typo error. Concerning the sentence: "when HIRA is overexpressed, PML NBs form in the absence of TNFb", it is inaccurate. Indeed, PML NBs are present in our cells with or without IFNb treatment. Overexpression of HIRA triggers accumulation of the ectopic HIRA in the PML NBs in absence of IFNb, probably as part of a buffering mechanism.

    Major concerns:

    • The suggested function of HIRA at the PML NBs as storage is interesting. Ideally, this would be tested by real-time single molecule tracking.

    While surely interesting, we believe that the real-time single molecule tracking is beyond the scope of our article. In addition, with our hypothesis that PML NBs act as buffering places for HIRA, HIRA might come in and out of PML NBs depending on its concentration and/or the availability of free binding sites and single molecule tracking might not be informative for long- term possible storage functions of PML NBs.

    • The link between PML NBs containing HIRA and H3.3 deposition is very intriguing and indeed the ChIP-seq data shown in Figure 5B shows a clear increase in the H3.3 signal around the TES. This distribution is very intriguing as recent work (Fang et al 2018 Nat Comm) showed that H3.3 deposition across the gene body was diverse and dynamic. Ideally, the qPCR of some select ISGs would confirm the ChIP-seq data. Here a more complex picture emerges. Just as with the ChIP-seq, a modest decrease of H3.3 at the TSS was observed, but only in 2 of the 3 genes shown is H3.3 enriched at the TES and only in 1 gene (ISG54) is H3.3 enriched at the gene body. As qPCR is later used in the manuscript (Fig 5E and 5G), it is essential that the results of two different techniques give similar results. With regards to Fig 5E and 5G, it is unclear why certain gene regions are shown, but not others.

    We agree with the reviewer that distribution of H3.3 on active genes follows a diverse and dynamic pattern. H3.3 is enriched on gene bodies but several papers have shown an important increase of H3.3 loading on the TES region of actively transcribed genes (Tamura et al. 2009; Sarai et al. 2013). Our ChIP-qPCR data (Figure 6A) and our ChIP-Seq data (Figure 6B) are consistent and show a moderate increase of H3.3 on gene bodies, eg on MX1 mid or ISG54 mid regions shown by qPCR on Figure 6A (this enrichment is reproducible but not necessarily statistically significant) and on gene bodies of the 48 core ISGs as shown in our ChIP-Seq data (see the light blue line between TSS and TES on figure 6B). In addition, our ChIP-qPCR and ChIP-Seq data also consistently show a higher enrichment of H3.3 on the TES regions of ISGs (see the significant enrichment found in ChIP-qPCR in the TES regions of MX1, OAS1 and ISG54, as well as the strong increase in H3.3 deposition with IFN seen by the light blue line for ChIP- Seq data on figure 6B).

    Since the strongest enrichment for H3.3 was found on the TES region, we focused on this region to evaluate the impact of HIRA or PML knock-down. Our ChIP-Seq data (now added in main Figure 6F for the whole ISG region, or with a zoom on the TES region in Figure 6G) shows that the strongest effect of HIRA or PML knock-down is indeed visible in the TES region of ISGs. Our ChIP-qPCR presented on Figure 6E data totally supports this effect.

    Overall, the link between HIRA and PML in H3.3 loading is only mildly affected (Fig 5E and 5F). The conclusion that HIRA and PML are essential (Page 12, line 8) is not represented by the presented data. The authors propose that DAXX could play a role. Indeed, work on another H3 variant, CENP-A, showed that non-centromeric localization is dependent on both HIRA and DAXX (Nye et al 2018 PLoS ONE). It would be interesting to learn if a double knock-down of HIRA and DAXX can prevent the enrichment of H3.3 at TES of ISGs upon TNFb treatment.

    To address the first part of the comment, we have now added 3 things :

    (1) we have tuned-down our conclusion by saying that HIRA and PML are 'important' for the long-lasting deposition of H3.3 on ISGs,

    (2) we provide new data of time-ChIP qPCR experiments suggesting that HIRA is important for H3.3 recycling during transcription of ISGs. We believe that these results strengthen the importance of HIRA for the global H3.3 enrichment on ISGs (by acting both in the de novo deposition and/or recycling of H3.3).

    We agree with the reviewer that it could be interesting to study the impact of the double knock-down of DAXX and HIRA on H3.3 enrichment at ISGs. However, we decided to focus our attention on SP100 since it could help us to better tease apart the role of HIRA localization in PML NBs, versus its role in H3.3 deposition at ISGs. In addition, since SP100 knock-down unleashes ISGs transcription, it also provided us with the opportunity to study the impact of an elevated ISGs transcription on H3.3 deposition and whether this is also mediated by HIRA.

    (3) we thus now also provide data of the double knock-down of SP100 and HIRA showing that the increase in H3.3 loading on ISGs seen upon SP100 knock-down is mediated by HIRA. This new result also strengthens the importance of HIRA for H3.3 enrichment on ISGs upon transcription.

    • In Figure 6B, two versions of HIRA are overexpressed and the authors conclude that the number of PML NBs goes up. Earlier in the manuscript, the authors showed that PML NB formation upon IFNb exposure brings HIRA into the PML NBs via a SUMO-dependent mechanism. Is overexpression of HIRA and its accumulation in PML NBs also SUMO-dependent or SUMO-independent? Overexpressing the SIM mutants from Figure 3F would address this question. In addition, the link between the proposed HIRA being stored at PML NBs could be strengthened by overexpressing HIRA and see at both short and late time points whether H3.3 is enriched on ISG genes.

    We want to clarify the first point: we do not conclude that the number of PML NBs goes up upon overexpression of HIRA. The number of PML NBs seems stable, although we have not quantified it. The aim of Figure 4A (previously Figure 6B) is to show that upon overexpression, ectopic forms of HIRA localize in PML NBs without IFN-I treatment, as part of a buffering mechanism.

    The SIM mutant of HIRA from Figure 3F is indeed overexpressed and does not localize in PML NBs upon IFN-I treatment. We have now added an IF (Figure 3- figure supplement 1C) showing that it does not localize either in PML NBs in non-treated cells. Thus, this underscores that accumulation of ectopic HIRA in PML NBs is SUMO-SIM-dependent regardless of the IFN-I treatment.

    • BJ cells are known to senesce rather easily. Did the authors double-check what fraction of their cells were in senescence and whether this correlated with the high or low expression of ectopic H3.3?

    BJ cells can indeed enter into senescence, but there are less prone to senesce than other human primary cells such as IMR90 for example. Nevertheless, we checked EdU incorporation both in BJ cells (Figure 1 - Figure supplement 1F) and BJ eH3.3i cells with expression of ectopic H3.3, with or without IFN-I treatment (Figure R2 for reviewer). We could clearly see that in our conditions (Dox addition for 24h maximum, IFNb at 1000U/mL for 24h), there is no significant difference in the number of EdU+ cells (ie proliferating cells), thus excluding effects due to senescence entry. As positive control, we have treated BJ cells with etoposide, a known senescence-inducing drug (Kosar et al., 2013; Tasdemir et al., 2016) which indeed reduces the number of EdU positive cells. We have now added a sentence in the main text as well to underscore that cells are not senescent.

    • In Figure 6 - figure supplement D, it appears that the levels of HIRA go up upon TSA and IFNb treatment. Rather than relying on visual inspection, ideally, all Western blots should be quantified to confirm the assessment that protein levels are not affected by different experimental procedures.

    We now provide quantification of all WBs below each WB. In addition, we have removed data on TSA since it could appear too preliminary.

    Reviewer #2 (Public Review):

    HIRA chaperone complex has been previously shown to localize at PML Nuclear Bodies upon various stress or stimuli (senescence, viral infections, interferon treatment). The authors have previously unraveled an anti-viral role of PML NBs through the chromatinization of HSV-1 viral genome by H3.3 chaperones. Here, the authors identify SUMOylation, as well as a SIM-like sequence in HIRA, as drivers for HIRA recruitment at PML Nuclear Bodies upon interferon-I treatment. These HIRA-containing PML NBs localize close to interferon-stimulated gene (ISG) loci. Although the functional role of HIRA/PML interaction is yet not solved, HIRA and PML regulate H3.3 loading at transcriptional end sites of IGS upon Interferon-I treatment. The authors propose that PML NBs play a buffering role for HIRA, regulating its availability depending on H3.3 level or chromatin dynamics.

    Strength:

    The authors used primary human diploid BJ fibroblasts, a relevant cell line for studying physiological regulation upon inflammatory cytokines. The role of SUMO/SIM on HIRA localization upon interferon beta treatment was assessed using interesting - but already described - tools, such as SUMO-specific affimers. The authors provide convincing results on the requirement of PML SUMOylation and a putative SIM sequence on HIRA for its localization at PML Nuclear Bodies. Other interesting observations are described, such as some PML or HIRA-dependent long-lasting H3.3 loading at transcription end site of ISGs upon interferon beta treatment, as shown by ChIP analyses of ISG loci, but also by endogenous H3.3 ChIPseq analysis.

    Weakness:

    The authors claim HIRA partitioning at PML NBs correlates with increase in "PML valency" upon interferon-I. The "valency" refers to the number of interaction domains, but the number of SUMOs conjugated on PML is not explored here (nor the number of SIMs on HIRA). Although the authors have proposed interested hypothesis and discussion, the inhibitory role of H3.3 overexpression or acetylation inhibition on HIRA localization at PML Nuclear Bodies clearly deserves further investigations.

    More generally, the manuscript explores many directions, but the links between the various observations remain unclear and limit firm conclusions.

    We thank the reviewer for his/her constructive comments.

    We have now addressed these 3 weaknesses pointed out by the reviewer.

    • Our claims on PML valency have been removed. We have now underscored the link between HIRA accumulation in PML NBs and the increase in PML and SP100 protein levels, without lingering on the valency aspects which was not the focus of our paper.

    • The role of H3.3 overexpression in inhibition of HIRA localization in PML NBs has been moved in the first part of the paper describing the mechanistic for accumulation of HIRA in PML NBs. We feel that these data are still of importance and support the role of PML NBs as a buffering place for HIRA depending on DAXX levels (new data) as well as H3.3 levels.

    We agree that the acetylation inhibition would deserve further investigations and we have thus removed the part on TSA treatment.

    • Thanks to the reviewer's comments, we have now remodeled the article to better convey two main conclusions : (1) PML NBs serve as a buffering site for HIRA. Accumulation of HIRA in PML NBs depends both on PML and SP100 concentration (and on PML SUMOylation) as well as DAXX and H3.3 levels and (2) upon IFN-I treatment, PML regulates ISGs transcription and thus indirectly regulates HIRA loading on ISGs, which controls H3.3 deposition and recycling during transcription. HIRA-mediated H3.3 deposition/recycling is highly dependent on ISGs transcription levels and is thus increased upon SP100 knock-down which unleashes ISGs transcription.
  3. eLife

    Evaluation Summary:

    In this manuscript, of interest to the fields of animal immunity and epigenetics, the authors investigate the crosstalk between PML Nuclear Bodies and HIRA, a member of the H3.3 histone chaperone complex, during inflammatory stress. This study raises interesting perspectives on how availability of HIRA could be regulated by PML Nuclear Bodies for histone deposition onto interferon-stimulated genes, which in turn, could be relevant for immune-response mediated gene regulation.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. The reviewers remained anonymous to the authors.)

  4. eLife

    Reviewer #1 (Public Review):

    The manuscript "Interplay between PML NBs and HIRA for H3.3 dynamics following type I interferon stimulus" by Kleijwegt and colleagues describes a study that's set out to explore the details of the PML-HIRA axis in H3.3 deposition at ISGs upon IFN-I stimulation. First, the authors establish that HIRA colocalized at PML NBs upon TNFa and TNFb treatment. This process is SUMO-dependent and facilitated by at least one of the identified SIM domains of HIRA. Next, the authors set out to determine whether interferon responsive genes (ISGs) are dependent on HIRA or PML. By knocking-down either HIRA or PML, only an effect on ISGs was observed when PML was knocked down. In fact, immune-FISH showed that PML NBs are in close proximity of ISGs upon TNFb treatment. To address the histone chaperone function of HIRA, the deposition of the replication-independent H3.3 on ISGs is tested. In specific, the enrichment of H3.3 across the ISG gene body. ChIP-seq data (Fig 5B) showed an enrichment around the TES, whereas qPCR (Fig 5A) showed less convincing enrichment (for details see below). When either HIRA or PML are knocked down, a mild loss of H3.3 enrichment was observed (Fig 5E). Interestingly, when HIRA is sequestered away from PML NBs by Sp100, an increased enrichment of H3.3 was observed. To understand the interplay between H3.3 deposition and HIRA's role in this process in the presence of PML NBs, H3.3 was overexpressed. Two population of cells were observed: low or high levels of H3.3. In the former, HIRA formed foci and the latter, HIRA did not form foci. Surprisingly, when HIRA is overexpressed, PML NBs form in the absence of TNFb. Finally, a two-sided model is proposed, where PML NBs is required for ISG transcription promoting H3.3 loading. The second side is that PML NBs function as a "storage center" for HIRA to regulate its availability.

    Overall, it the model is intriguing, but the data presented seems insufficient to support the current claims.

    Major concerns:
    - The suggested function of HIRA at the PML NBs as storage is interesting. Ideally, this would be tested by real-time single molecule tracking.

    - The link between PML NBs containing HIRA and H3.3 deposition is very intriguing and indeed the ChIP-seq data shown in Figure 5B shows a clear increase in the H3.3 signal around the TES. This distribution is very intriguing as recent work (Fang et al 2018 Nat Comm) showed that H3.3 deposition across the gene body was diverse and dynamic. Ideally, the qPCR of some select ISGs would confirm the ChIP-seq data. Here a more complex picture emerges. Just as with the ChIP-seq, a modest decrease of H3.3 at the TSS was observed, but only in 2 of the 3 genes shown is H3.3 enriched at the TES and only in 1 gene (ISG54) is H3.3 enriched at the gene body. As qPCR is later used in the manuscript (Fig 5E and 5G), it is essential that the results of two different techniques give similar results. With regards to Fig 5E and 5G, it is unclear why certain gene regions are shown, but not others.
    Overall, the link between HIRA and PML in H3.3 loading is only mildly affected (Fig 5E and 5F). The conclusion that HIRA and PML are essential (Page 12, line 8) is not represented by the presented data. The authors propose that DAXX could play a role. Indeed, work on another H3 variant, CENP-A, showed that non-centromeric localization is dependent on both HIRA and DAXX (Nye et al 2018 PLoS ONE). It would be interesting to learn if a double knock-down of HIRA and DAXX can prevent the enrichment of H3.3 at TES of ISGs upon TNFb treatment.

    - In Figure 6B, two versions of HIRA are overexpressed and the authors conclude that the number of PML NBs goes up. Earlier in the manuscript, the authors showed that PML NB formation upon IFNb exposure brings HIRA into the PML NBs via a SUMO-dependent mechanism. Is overexpression of HIRA and its accumulation in PML NBs also SUMO-dependent or SUMO-independent? Overexpressing the SIM mutants from Figure 3F would address this question. In addition, the link between the proposed HIRA being stored at PML NBs could be strengthened by overexpressing HIRA and see at both short and late time points whether H3.3 is enriched on ISG genes.

    - BJ cells are known to senesce rather easily. Did the authors double-check what fraction of their cells were in senescence and whether this correlated with the high or low expression of ectopic H3.3?

    - In Figure 6 - figure supplement D, it appears that the levels of HIRA go up upon TSA and IFNb treatment. Rather than relying on visual inspection, ideally, all Western blots should be quantified to confirm the assessment that protein levels are not affected by different experimental procedures.

  5. eLife

    Reviewer #2 (Public Review):

    HIRA chaperone complex has been previously shown to localize at PML Nuclear Bodies upon various stress or stimuli (senescence, viral infections, interferon treatment). The authors have previously unraveled an anti-viral role of PML NBs through the chromatinization of HSV-1 viral genome by H3.3 chaperones. Here, the authors identify SUMOylation, as well as a SIM-like sequence in HIRA, as drivers for HIRA recruitment at PML Nuclear Bodies upon interferon-I treatment. These HIRA-containing PML NBs localize close to interferon-stimulated gene (ISG) loci. Although the functional role of HIRA/PML interaction is yet not solved, HIRA and PML regulate H3.3 loading at transcriptional end sites of IGS upon Interferon-I treatment. The authors propose that PML NBs play a buffering role for HIRA, regulating its availability depending on H3.3 level or chromatin dynamics.

    Strength:
    The authors used primary human diploid BJ fibroblasts, a relevant cell line for studying physiological regulation upon inflammatory cytokines. The role of SUMO/SIM on HIRA localization upon interferon beta treatment was assessed using interesting - but already described - tools, such as SUMO-specific affimers. The authors provide convincing results on the requirement of PML SUMOylation and a putative SIM sequence on HIRA for its localization at PML Nuclear Bodies. Other interesting observations are described, such as some PML or HIRA-dependent long-lasting H3.3 loading at transcription end site of ISGs upon interferon beta treatment, as shown by ChIP analyses of ISG loci, but also by endogenous H3.3 ChIPseq analysis.

    Weakness:
    The authors claim HIRA partitioning at PML NBs correlates with increase in "PML valency" upon interferon-I. The "valency" refers to the number of interaction domains, but the number of SUMOs conjugated on PML is not explored here (nor the number of SIMs on HIRA).
    Although the authors have proposed interested hypothesis and discussion, the inhibitory role of H3.3 overexpression or acetylation inhibition on HIRA localization at PML Nuclear Bodies clearly deserves further investigations.
    More generally, the manuscript explores many directions, but the links between the various observations remain unclear and limit firm conclusions.

  6. Version published to 10.1101/2021.11.30.470516v2 on bioRxiv
  7. Version published to 10.1101/2021.11.30.470516v1 on bioRxiv