Live-cell 3D single-molecule tracking reveals how NuRD modulates enhancer dynamics

S Basu, O Shukron, A Ponjavic, P Parruto, W Boucher, W Zhang, N Reynolds, D Lando, D Shah, LH Sober, A Jartseva, R Ragheb, J Cramard, R Floyd, G Brown, K Gor, J Balmer, TA Drury, AR Carr, L-M Needham, A Aubert, G Communie, L Morey, E Blanco, MA Barber, I Mohorianu, T Bartke, L Di Croce, I Berger, C Schaffitzel, SF Lee, TJ Stevens, D Klenerman, BD Hendrich, D Holcman, ED Laue

Preprint posted on 4 April 2020


Live-cell single particle tracking of PRC1 reveals a highly dynamic system with low target site occupancy

Miles K. Huseyin, Robert J. Klose

Preprint posted on 25 April 2020

Article now published in Nature Communications at

Tracking chromatin complexes at the single-molecule level: NuRD and PRC1 in the limelight

Selected by Gabriel Aughey

preLight written together with Mate Palfy


Precise control of gene expression is largely mediated through the actions of chromatin-associated protein complexes. These complexes can have a variety of biochemical activities that result in nucleosome remodelling, posttranslational modification of histones, or the recruitment of further proteins. Despite extensive research into the enzymatic activities of these complexes and characterisation of their genome-wide binding distributions on chromatin, the mechanisms by which they coordinate changes in gene expression are still poorly understood. Recent advances in imaging now make it possible to track individual molecules of transcriptional regulators in living cells to better understand their behavior. Using similar in vivo single-molecule imaging methods, two recent preprints dissect how the chromatin regulators NuRD (Nucleosome Remodelling and Deacetylation) and PRC1 (Polycomb Repressive Complex 1) bind and modify chromatin, providing new insights into how they regulate transcription.


Key findings Basu et al.

In the first preprint the authors examine the Nucleosome Remodelling and Deacetylation (NuRD) complex. This conserved multi-protein complex is required for the maintenance of pluripotency and regulation of cell-fate transitions. Its two main biochemical activities are mediated by the ATP-dependent nucleosome remodelling subunit (CHD4) and the histone deacetylase subunit (HDAC), but how the complex as a whole regulates gene expression is unclear.

Having observed from Hi-C experiments that cells in which NuRD activity is disrupted display a loss of intermediate-range contacts, the authors use live-cell single-molecule tracking to better understand the impact of NuRD on chromatin architecture and enhancer-promoter dynamics.
By imaging Halo-tagged CHD4 in mouse ESCs, they observe NuRD complexes in two states: a fast, freely diffusing state and a slow state representing chromatin-bound complexes.The authors next turned their attention to the behaviours of the nucleosome remodelling CHD4 subunit and the HDAC subunit. By knocking down MBD3, which is thought to act as a bridge between the two subcomplexes, the authors were able to observe the dynamics of HDAC recruitment. In MBD3 knockdown cells the binding of the HDAC-containing subcomplex to chromatin was significantly reduced, suggesting that the HDAC subcomplex has a low affinity for DNA and that MBD3 is required for assembly of the holocomplex.

Careful analysis of the movement of chromatin bound CHD4 showed decreased measures of NuRD movement when bound to DNA in MBD3 knockdown cells. This implies that NuRD-bound chromatin is more condensed when the HDAC subcomplex is absent.

To further assess the impact of NuRD binding on 3D chromatin structure, the authors made use of the CARGO-dCas9 system which labels specific sequences by using gRNAs to recruit catalytically inactive fluorescently tagged Cas9 proteins. By tracking the movement of the fluorescent signal from Cas9, the dynamics of targeted chromatin loci can be analysed. These experiments indicated that the movement of the chromatin itself was restricted in MBD3 knockdown cells, rather than just the movement of the NuRD complex. These data are surprising since the presence of HDACs is usually assumed to be associated with compacted chromatin.

From these data, Basu et al., propose a model in which NuRD association with chromatin results in local decompaction which increases the search area available for enhancer-promoter contacts to occur.


Key findings Huseyin & Klose

 In the second preprint, the authors interrogate the binding behaviour of Polycomb Repressive Complex 1 (PRC1), which is one of the best-studied complexes when it comes to the regulation of gene repression. PRC1 is an E3 ubiquitin ligase that deposits the posttranslational modification ubiquitin on histone H2A at lysine position 119 (H2AK119ub1). However, whether this specific modification (and thus PRC1’s catalytic activity) is required for gene repression, or whether PRC1 binding affects transcription through chromatin accessibility, chromatin compaction, phase separation or other non-catalytic ways is highly debated. A quantitative characterisation of how PRC1 binds its target in vivo could help distinguish which model(s) best explains PRC1’s mechanism of action – which is what this study set out to achieve.

By endogenously tagging the core subunit of canonical PRC1 complexes (RING1B) and performing single-particle tracking, the authors first establish that a 3-state kinetic model of immobile, slowly diffusing and fast diffusing molecules best fits their data. The first surprising finding they make is that only 20% of PRC1 molecules is found in the immobile chromatin-bound state, with only a fraction (38%) of those binding events being stable. Combining these experiments with a biochemical approach and available Ring1B ChIP-seq data, their quantification shows that there are 0.1 – 0.3 molecules per kilobase of RING-1B enriched chromatin. Such sparsity of PRC1 at its target sites makes a non-catalytic model of PRC1-mediated repression difficult to reconcile.

The authors also perform similar measurements for the canonical PRC1 component PCGF2, and variant PRC1 complexes containing PCGF1, PCGF3, PCGF5 and PCGF6 components.  Canonical PRC1-PCGF2 showed stable binding, while interestingly the variant complexes had a range of binding characteristics. For example, PCGF3 which has a role in pervasive ubiquitylation of the genome, displayed smaller bound fractions and shorter binding times compared to the canonical complex, while the variant PCGF1 displayed the highest chromatin-bound fraction (26%).

What factors regulate how frequently and how stably PRC1 binds its target sites? H3K27me3, which is deposited by PRC2 (the second main Polycomb complex), is known to facilitate PRC1 binding; using PRC2 core subunit mutant cells the authors show that H3K27me3 recognition increases the frequency, but not stability of PRC1 binding. In contrast, a specific domain of the variant PCGF1 which interacts with DNA-binding proteins impacts variant PCGF1-PRC1 binding stability, but not binding frequency.

Taken together, this work provides an in-depth quantitative analysis of PRC1 binding that supports the model that PRC1 regulates gene expression through its E3 catalytic activity. The authors also point to key mechanisms that influence the binding behaviour of PRC1.


Similarities and differences

Whilst the authors of these studies have examined two different complexes with distinct cellular phenotypes, this pair of preprints provide a clear demonstration of the power of single-molecule tracking to gain insights into the regulation of gene expression by multi-protein chromatin-modifying complexes.

Both NuRD and PRC1 possess enzymatic components that mediate catalytic modification of histones (HDAC and E3 ubiquitin ligase, respectively). However, the contribution of these reactions to the complexes’ functions are not always clear since the complex as a whole may act to physically regulate chromatin through other means. Using single-molecule microscopy, these two studies reach relatively different conclusions about the importance of catalysis in NuRD or PRC binding. Whilst Huseyin & Klose propose that catalysis with transient chromatin association is the dominant mode of action of PRC1, Basu et al find that the long-lasting binding of the complex to target loci mediates NuRD action with HDAC imparting a surprising chromatin decompacting function.

In terms of techniques, both Basu et al. and Huseyin & Klose made use of Halo-tagged proteins for imaging. Although the use of Halo-tags with their cognate dyes results in incredibly bright fluorophores with good photostability, both preprints noted that photobleaching still occurred when attempting to track stable chromatin bound complexes over relatively long timescales. These limitations highlight the need for development of even more photostable dyes to gain further insights into potential mechanisms by which stably bound complexes regulate chromatin over long timescales.


Posted on: 22 May 2020


(No Ratings Yet)

Author's response

Basu et al. shared about Live-cell 3D single-molecule tracking reveals how NuRD modulates enhancer dynamics

How can you account for the surprising observation that HDAC recruitment appears to be associated with decondensed chromatin? Is it possible that these effects of the HDAC containing subcomplex are associated with another biochemical function independent of histone acetylation? 

Loss of Mbd3 causes dissociation of the Chd4 remodelling and HDAC subunits within NuRD.  We show that Chd4 appears to be the main determinant of chromatin binding within NuRD, so loss of Mbd3 is expected to result in failure or reduction of HDAC subunit recruitment to Chd4-bound sites.  It is indeed surprising that this loss results in increased movement of bound Chd4 and of chromatin.

How can we explain this surprising finding?  It is important to remember that restoration of Mbd3 in ES cells results in a change in Chd4’s nucleosome remodelling activity: although we could detect only a minimal change in Chd4 abundance, and no change in its bound sites, we saw a rapid movement of nucleosomes on reassembly of the intact NuRD complex (Bornelöv, Reynolds et al. 2018).  Whilst in vitro studies suggest that Chd4’s nucleosome remodelling activity does not change upon addition of other NuRD components (Zhang et al. 2016; Zhong et al. 2020), inside cells we thus find that the loss of MBD3 causes a change in Chd4’s remodelling activity suggesting that this might result in the change in chromatin condensation that we see.  This result underlines the importance of studying protein complexes in the context of chromatin and the other complexes in which they operate.


You showed that intact NuRD leads to an increase of CTCF and cohesin binding. Do you think that NuRD directly interacts with CTCF and/or cohesin to stablise the chromatin environment, or do you think it is more likely that remodelling of chromatin by NuRD creates a chromatin environment that is amenable to CTCF/cohesin binding?

Although we do not know, the second option may be more likely because we have not been able to find any evidence for direct interactions between NuRD and CTCF/cohesin.  However, because cohesin can slide along DNA and is known to have complex cycles of loading and unloading, there could be several possible ways in which NuRD could impact CTCF/cohesin complex formation.  It will be interesting to explore whether remodellers either block or facilitate the movement of cohesin along the DNA to specific sites, or whether they somehow affect the process of cohesin loading/unloading.



Bornelöv S, Reynolds N, Xenophontos M, Gharbi S, Johnstone E, Floyd R, Ralser M, Signolet J, Loos R, Dietmann S, Bertone P, Hendrich B.  The Nucleosome Remodeling and Deacetylation Complex Modulates Chromatin Structure at Sites of Active Transcription to Fine-Tune Gene Expression.  Mol Cell. 2018 Jul 5;71(1):56-72.e4.


Zhang W, Aubert A, Gomez de Segura JM, Karuppasamy M, Basu S, Murthy AS, Diamante A, Drury TA, Balmer J, Cramard J, Watson AA, Lando D, Lee SF, Palayret M, Kloet SL, Smits AH, Deery MJ, Vermeulen M, Hendrich B, Klenerman D, Schaffitzel C, Berger I, Laue ED.  The Nucleosome Remodeling and Deacetylase Complex NuRD Is Built from Preformed Catalytically Active Sub-modules.  J Mol Biol. 2016 Jul 17;428(14):2931-42.


Zhong Y, Paudel BP, Ryan DP, Low JKK, Franck C, Patel K, Bedward MJ, Torrado M, Payne RJ, van Oijen AM, Mackay JP.  CHD4 slides nucleosomes by decoupling entry- and exit-side DNA translocation.  Nat Commun. 2020 Mar 23;11(1):1519.


Huseyin and Klose shared about Live-cell single particle tracking of PRC1 reveals a highly dynamic system with low target site occupancy

Your results suggest that PRC1 represses genes through its catalytic activity in ESCs. Do you think that during developmental transitions where gene expression programs need to be robustly activated or repressed, non-catalytic functions of PRC1 might be also important for repression at Polycomb-regulated genes? For example at Polycomb bodies, which are developmentally regulated and where you observe more stable long-term binding?

It is likely that the relative contribution of catalytic and non-catalytic PRC1 activities to PRC1-depedent gene regulation will vary during development. In rapidly dividing embryonic stem cells, PRC1-dependent gene repression appears to rely primarily on catalysis and H2AK119ub1. However, as the cell cycle slows in more committed cell types, or cells terminally differentiate, non-catalytic activities that create structural effects on chromatin may have more central roles in gene repression. These mechanisms could, for example, function to increase the fidelity of catalysis-dependent repression mechanisms. Indeed, previous work in developmental model systems has provided evidence to this effect. In addition, how PRC1 is used to achieve appropriate gene regulation may depend on the composition of PRC1 complexes in specific cell types. For example, the biochemical make-up of canonical PRC1 complexes, which have catalysis-independent effects on gene expression and are necessary for Polycomb body formation, change as cells differentiate. In embryonic stem cells, canonical PRC1 predominantly incorporates molecules that do not appear to exhibit phase separating behaviour. However, as cells differentiate, the composition of these complexes changes to include molecules that have the capacity to phase separate, possibly altering the repressive capacity of canonical PRC1. Clearly, the complexity of PRC1 means we still have a lot to learn about the breadth of its repressive mechanisms and how these are used to control gene expression in different cell types and developmental contexts.


Related to this, you observe a small fraction of very stable PRC1 binding (over 100 s) in ESCs; do you think this fraction could increase during differentiation? Do you have plans for follow-up studies to image PRC1 binding dynamics while differentiating ESCs, and is this feasible to do?

The very stably bound fraction of PRC1 could possibly increase during differentiation. This could be achieved if the proportion of canonical PRC1 complexes that display very stable binding events were to increase during these cellular transitions. Alternatively, as the composition of PRC1 complexes changes during differentiation, and new auxiliary subunits are incorporated, PRC1 complexes could acquire distinct chromatin binding behaviours. Given the tools and approaches we have developed in the context of this work, it will be feasible to address these and other important questions in future studies, and this is something we look forward to doing. However, perhaps most importantly, major advances in live-cell imaging approaches over the past decade have opened up new avenues to study the inner workings of the cell. Being able to watch and measure the behaviours of individual molecules inside live cells has created new opportunities to understand Polycomb system function, and studies like ours have shown that these approaches provide essential new insight into the function of these fascinating chromatin-based gene regulatory systems.

Have your say

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Sign up to customise the site to your preferences and to receive alerts

Register here