Menu

Close

The modular mechanism of chromocenter formation in Drosophila

Madhav Jagannathan, Ryan Cummings, Yukiko M Yamashita

Preprint posted on November 28, 2018 https://www.biorxiv.org/content/early/2018/11/28/481820

Article now published in eLife at http://dx.doi.org/10.7554/elife.43938

Not just junk DNA: Satellite DNA and their cognate binding proteins form chromocenters to gather up chromosomes in one nucleus

Selected by Maiko Kitaoka

Background

Eukaryotic chromosomes commonly have regions of satellite DNA abundant in tandem repeats at centromeric and pericentromeric chromatin. Centromeric heterochromatin is well known to establish kinetochore function and allow for faithful chromosome segregation. However, the role of pericentric satellite DNA is less understood, particularly due to the lack of protein-coding and conservation across species. Previous studies have shown roles for specific contexts only, but whether there is a more central role for eukaryotic satellite DNA, especially given its abundance in the genome, is not known.

With this new preprint, Jagannathan, Cummings, and Yamashita follow up on their recent eLife study discussing the formation and role of chromocenters, or the bundling of multiple chromosomes, to maintain nuclear organization. They address several new questions, including how all Drosophila melanogaster chromosomes in a genome can be bundled in chromocenters, and how multiple satellite DNAs and their corresponding DNA binding proteins can modulate this process together.

Identification of a different satellite DNA sequence that labels the autosomes, rather than X and Y chromosomes. From Figure 1.

 

Key findings

Since their previous work identified a protein, D1, that was responsible for bundling chromosomes X, Y, and 4 into chromocenters, the group turned their attention to chromosomes 2 and 3, the major autosomes in D. melanogaster, to address how all chromosomes in the genome may be grouped into chromocenters. This led them to discover a longer satellite repeat on the autosomes that was bound by the Prod (proliferation disrupter) protein. Disrupting prod function led to micronuclei formation and increased DNA damage, ultimately resulting in cellular death. This was also seen in their previous work with D1 mutants. Interestingly, while D1 mutations affected the germline, prod mutations did not, thus suggesting that both act on different tissues in spite of their similar chromocenter-forming functions. However, D1 prod double mutants fail to develop past the embryonic stages and have increased micronuclei, establishing the essential requirement of chromocenter formation via satellite DNA.

Loss of prod led to formation of micronuclei in imaginal discs (left, A and B) and lymph glands (right, C and D). From Figure 2.

 

Prod clearly has a similar role to D1 in forming chromocenters. The authors expressed Prod ectopically in a tissue where it’s normally not present and caused the formation of chromatin threads that establish various chromosome territories connecting the autosomes. This clearly shows that Prod is sufficient to bundle these chromosomes together through its satellite DNA binding, explaining the mechanism of how chromocenters are formed.

At this point, Jagannathan and Cummings et al have established that Prod and D1 cluster chromosomes through satellite DNA. But since both proteins act on different sets of chromosomes, how is an entire genome’s set of chromosomes encapsulated in a nucleus? Prod and D1 did not seem to interact together through immunoprecipitation, suggesting only a weak or transient interaction. Through live imaging, the authors discovered a “kiss and run” interaction – D1 foci and Prod foci touch briefly and then separate, indicating a dynamic process of chromocenter formation. Both proteins seem to be mutually dependent on the other’s functional presence as well, since prod mutants showed defective D1 clustering in nuclei and vice versa. This interdependency provides a network to establish the bundling of all chromosomes in a genome.

Model showing how chromocenters are formed with different satellite DNA and binding proteins. From Figure 5.

 

In summary, Jagannathan and Cummings et al have demonstrated that Prod and D1 create a network where both proteins bind their respective satellite DNA sequences in order to bring all chromosomes into a chromocenter, and eventually package the entire genome properly into the nucleus. Their study demonstrates the importance of satellite DNA, addressing not only the molecular and cell biological consequences of Prod and D1 perturbations but also the evolutionary significance of these proteins and satellite DNA as well.

Questions for the authors

The elife study showed that D1 loss causes micronuclei to bud off from the main nucleus – does prod loss cause micronuclei formation in the same way?

Do Prod and D1 interact with other proteins apart from their transient “kiss and run” interactions? Are these proteins known to function in other processes related to nuclear integrity and/or genome packaging?

In organisms with more than 4 chromosomes, how many chromocenter-forming, satellite DNA-binding proteins might be necessary to bundle a larger genome with more chromosomes?

The discussion mentions that D. simulans doesn’t have the satellite DNA that binds D. melanogaster Prod, so what is the Prod-like protein in D. simulans and how similar is it to D. melanogaster in structure and function?

Tags: cell biology, drosophila, fruit flies, genetics, genome packaging, satellite dna

Posted on: 11th December 2018

Read preprint (No Ratings Yet)




  • Author's response

    The authors shared

    1) The elife study showed that D1 loss causes micronuclei to bud off from the main nucleus – does prod loss cause micronuclei formation in the same way?

    This is indeed exactly what we predict causes micronuclei in prod mutant tissues but have not been able to test this yet, largely for technical reasons.

    2) Do Prod and D1 interact with other proteins apart from their transient “kiss and run” interactions? Are these proteins known to function in other processes related to nuclear integrity and/or genome packaging?

    Both D1 and Prod are predicted to interact with multiple proteins based on a high throughput mass spec screen (DroID database) and this is how we identified the D1-Prod interaction in the first place. Given that Drosophila melanogaster contains at least 17 distinct satellite DNAs which are most likely present in chromocenters, we anticipate that D1 and Prod interact with multiple other satellite DNA binding proteins that have yet to be identified.

    D1 and Prod are well-known to be associated with constitutive heterochromatin, affecting phenotypes such as position effect variegation.  Given the major role played by both proteins in satellite DNA packaging within nuclei, it is currently difficult to discriminate between satellite DNA-dependent and satellite DNA-independent functions for D1 and Prod.

    3) In organisms with more than 4 chromosomes, how many chromocenter-forming, satellite DNA-binding proteins might be necessary to bundle a larger genome with more chromosomes?

    Although this is not common among eukaryotes, the mouse genome only contains a single pericentromeric satellite DNA repeat and we have shown that the HMGA1 protein bundles this satellite DNA into chromocenters in our previous eLife paper. Therefore, we think that it is the number of unique satellite DNA repeats rather than the number of chromosomes or genome size that dictates how many satellite DNA binding proteins are required to bundle a genome.

    4) The discussion mentions that D. simulans doesn’t have the satellite DNA that binds D. melanogaster Prod, so what is the Prod-like protein in D. simulans and how similar is it to D. melanogaster in structure and function?

    Interestingly, D. simulans does contain a Prod orthologue, which is highly conserved. Our idea is that D. simulans must have a satellite DNA-cognate binding protein pair that fulfill the same function as D. melanogaster Prod and the {AATAACATAG}n satellite DNA. Whether this involves the D. simulans Prod protein is unclear at the moment and we’re trying to come up with ways to tackle this question.

    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

    preLists in the cell biology category:

    Close