Menu

Close

The coordination of terminal differentiation and cell cycle exit is mediated through the regulation of chromatin accessibility

Yiqin Ma, Daniel J McKay, Laura Buttitta

Preprint posted on December 06, 2018 https://www.biorxiv.org/content/early/2018/12/06/488387

Chromatin accessibility on the wing – insights into chromatin regulation of cell cycle exit

Selected by Gabriel Aughey

 

Background

The development of complex animal tissues requires the finely tuned expression of thousands of regulatory genes. Changes to underlying chromatin landscape are required to mediate the changes in gene expression necessary for fully differentiated tissues to arise. Consequently, chromatin landscapes are dynamic during development. Whilst it is commonly accepted that accessible or “open” chromatin is associated with regulatory elements that control gene expression, the full relationship between chromatin accessibility and development is incompletely understood.

Much work has been done on understanding the developmental programs that drive differentiation, however the question of how development is brought to a halt at the chromatin level (i.e. terminal differentiation and cell cycle exit) is less well understood. In this preprint, Ma et al. investigate the chromatin changes that accompany wing differentiation in Drosophila and uncover some surprising findings regarding the open chromatin at cell-cycle gene loci.

 

Key findings

Chromatin accessibility undergoes extensive changes concurrent with altered gene expression during wing development

Ma et al. initially set out to characterise the gene expression and chromatin accessibility profiles of Drosophila wing tissue at various timepoints during development. To achieve this, they conducted RNA-seq and FAIRE-seq at 6 developmental timepoints in the larval and pupal stages (with the final stage comprising post-mitotic differentiated cells). Unsurprisingly, the authors discovered both highly dynamic gene expression and chromatin accessibility landscapes during wing differentiation. Gene expression was correlated to chromatin accessibility and the majority of accessibility changes were observed to be associated with enhancers becoming activated. Importantly, cell-cycle genes were seen to have dynamic regions of open chromatin that correlated with known cell-cycle changes in the developing wing. The critical cell-cycle regulators CycE, stg and e2f1 all exhibited regions in which loss of chromatin accessibility was observed at cell-cycle exit.

 

Disruption of cell cycle has little impact on chromatin accessibility of cell cycle genes

Having confirmed that chromatin accessibility is correlated with cell-cycle gene expression at cell-cycle exit, the authors sought to understand whether the observed closing of chromatin was a function of cell-cycle exit, or a precursor to it. To answer this question cell cycles were artificially extended through the overexpression of E2F alone or E2F with cyclinD, resulting in either one extra cell division or several rounds of mitoses, respectively. Intriguingly, when wing tissue was profiled by RNA-seq and FAIRE-seq under these conditions, surprisingly few changes to chromatin accessibility were observed, despite extensive changes to gene expression. When several critical cell cycle regulator gene loci were examined in more detail in ectopically cycling tissue (e.g. CycE, stg), chromatin accessibility at putative enhancers was almost indistinguishable from postmitotic cells at the same timepoint (i.e. enhancers continue to be rendered inaccessible despite continued expression of the gene – see figure). This unusual finding suggests that the closing of chromatin is regulated by developmental cues that are not directly linked to the cell-cycle.

 

Fig.1. Chromatin accessibility at key cell-cycle regulator cyclinE remains unchanged when cell-cycles are artificially extended despite continued expression of the gene (Figure 4E-D in preprint).

 

Cell cycle disruption impacts accessibility of chromatin at terminal differentiation genes

Whilst changes to chromatin accessibility were not observed at cell cycle genes, the authors went on to show that a subset of genes involved in terminal differentiation displayed a loss of open chromatin when cell cycle exit was delayed compared to controls. These regions largely represented enhancers that were seen to open following cell-cycle exit rather than open regions that became closed as a result of continued cell-cycling. In contrast to the cell cycle genes, this correlated with changes in expression and ultimately resulted in cuticle formation phenotypes.

 

Why I like this preprint

I found this preprint appealing due to the comprehensive approach the authors have taken to characterise gene expression and chromatin accessibility, resulting in a window into wing development with impressive temporal resolution.

The finding that chromatin accessibility and cell-cycle can be decoupled is surprising and seems to fly in the face of conventional thinking. This new perspective may force us to rethink our ideas about how terminal differentiation is achieved.

 

Questions for the authors

If chromatin accessibility changes occur independently of cell cycles and are not sufficient to disrupt expression (i.e. CycE is expressed highly in bypassed G0 despite loss of chromatin accessibility – Fig4D), then what is the purpose of the chromatin closing?

Is it possible that the key enhancers for the expression of cell-cycle genes are located distally and act at long range, so are difficult to detect? (i.e. the proximal open chromatin regions observed to be unaffected are redundant with long range enhancers?)

Do you expect that there are changes occurring to histone modifications at cell-cycle loci at which chromatin accessibility appears unaffected by continued cell-cycles?

Do you have any ideas about how the developmental closing of cell cycle genes (i.e. CycE, stg) chromatin is programmed if not directly linked to cell cycle, and do you expect that this will be a conserved mechanism across cell types?

 

Tags: chromatin, development, drosophila

Posted on: 7th January 2019

Read preprint (No Ratings Yet)




  • Author's response

    Laura, Yiqin and Dan shared

    If chromatin accessibility changes occur independently of cell cycles and are not sufficient to disrupt expression (i.e. CycE is expressed highly in bypassed G0 despite loss of chromatin accessibility – Fig4D), then what is the purpose of the chromatin closing?

     Our hypothesis is that the closing of chromatin at distal enhancers for key cell cycle regulators limits the ability of developmental signaling pathways to induce proliferation in the differentiating wing. For example, the closing enhancers that we verified drive expression of cycE and stg in the pupa wing, contain known binding sites for the Notch effector Su(H) and Yorkie/Scalloped (Yki/Sd), the downstream effector inhibited by Hippo signaling. We propose that closing of these enhancers allows pathways such as Notch and Hippo to be re-used to perform other postmitotic functions in the late wing without reactivating cell cycle genes or proliferation. Indeed, we know from our previous studies that Notch signaling and even Yorkie activation are unable to promote continued proliferation in the late pupa wing, despite promoting extensive proliferation at earlier stages. We believe the closing of the enhancers containing SuH and Yki/Sd binding sites at cycE and stg, at least in part, explains this puzzling result.

     

    In our experiments to compromise cell cycle exit, we are using very specific combinations of cell cycle regulators such as direct overexpression of E2F and CycD/Cdk4. We use these combinations because they are the only tools we know of that can prevent cell cycle exit at late stages without preventing terminal differentiation (and we have tested many, many pathways and combinations over the years). In these experiments we believe we are bypassing the distal enhancers and re-activating expression through promoter proximal regulatory elements, that remain accessible, but may not be the normal enhancers used in the pupa wing. It is worth noting there may also be a role for the number of enhancers that close in determining the robustness of silencing during cell cycle exit. For example, cycE only has one closing enhancer and can be readily reactivated by ectopic E2F activity postmitotically, while stg has multiple enhancers closing and requires additional CycD/Cdk4 activity to overcome the barrier to re-expression.

     

    Is it possible that the key enhancers for the expression of cell-cycle genes are located distally and act at long range, so are difficult to detect? (i.e. the proximal open chromatin regions observed to be unaffected are redundant with long range enhancers?)

    Absolutely this could be true. In the preprint we validate that some of the putative enhancers are sufficient to drive expression in the pupa wing, but we have many, many more to test and none of the enhancers we have examined thus far are sufficient to fully recapitulate endogenous expression. There is a lot of work ahead on this project. We would like to make overlapping deletion constructs and potentially use CRISPR to test redundancy of some of these putative regulatory elements. Now that we have a map of the potential pupa wing enhancers we are finally in a position to begin figuring out their logic.

     

    Do you expect that there are changes occurring to histone modifications at cell-cycle loci at which chromatin accessibility appears unaffected by continued cell-cycles?

    Yes. For many of the downstream cell cycle genes with simple enhancers that do not exhibit chromatin closing, we expect that the accessible regions proximal to the promoter bind the dREAM (RB/E2F) repressive complex leading to repressive histone modifications. When we disrupt this complex by flooding the system with ectopic E2F, this leads to activator E2F complexes binding in the same accessible regions, and subsequent recruitment of histone modifiers that help activate expression.

     

    Do you have any ideas about how the developmental closing of cell cycle genes (i.e. CycE, stg) chromatin is programmed if not directly linked to cell cycle, and do you expect that this will be a conserved mechanism across cell types?

    Our hypothesis is that the major developmental cue driving cell cycle exit and chromatin closing is ecdysone signaling, either directly or indirectly. Our previous work has shown a clear role for ecdysone signaling in promoting cell cycle exit in the wing and we have shown that ecdysone targets such as the transcription factor E93 dramatically remodel chromatin accessibility in the pupa wing.

    Based upon our preliminary data, we expect a similar pathway is at play in multiple tissues and cell types that exhibit robust cell cycle exit at the same timepoints we examined in the pupa. These include certain cell types in the eye and brain, as well as the legs and haltere. Importantly, while ecdysone seems to promote cell cycle exit in these cell types in the pupa, other cell types such as the abdominal histoblasts respond to ecdysone quite differently, and instead trigger proliferation during pupa stages. We therefore think there are common mechanisms at play in a subset of cell types, but there must be additional mechanisms at play that can alter the response to the same hormone differently in cells that proliferate in the pupa.  We believe the way forward to resolve this conundrum is to figure out the identity of the chromatin remodeler involved in chromatin closing at cell cycle genes, and how it may be regulated in a tissue-specific and cell type specific manner by ecdysone signaling.

    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
    Close