Research

Click here for: Lay readers' summary of our research

Click here for: Research 2020-2022 and Research 2011-2020 

Chromosome compaction is essential for chromosome segregation during cell division and regulate 3D genome organization and gene expression during interphase. The goal of my lab is to understand the molecular mechanisms that regulate chromosome structure and function. A key feature of eukaryotic chromosomes is DNA loops, formed by the structural maintenance of chromosomes (SMC) family of proteins. Two SMC complexes condensin and cohesin are ATP dependent molecular motors that bind to and translocate on DNA forming progressively enlarging loops through a process named “DNA loop extrusion”. In all organisms condensin-mediated loop extrusion compacts mitotic chromosomes in preparation for cell division. In mammalian cells, cohesin-mediated loop extrusion forms topologically associating domains (TAD) that regulate enhancer-promoter specificity. Current projects in the lab address the evolution and mechanisms of condensin and cohesin mediated genome organization and gene regulation using Caenorhabditis elegans and other nematode species.  

3D organization of the C. elegans genome 

C. elegans genome is partitioned into six similar sized chromosomes, including five autosomes (I-V) and one X chromosome (Figure 1). The chromosomes are holocentric (multiple centromeres). The central regions are gene rich and make inter chromosomal contacts. Left and right arms are repeat rich and are located close to the nuclear lamina (Ikegami et al. 2010, Sawh et al. 2020, Bian et al. 2020). At the mega-base scale, the autosomes display A/B compartmentalization without loop-anchored TADs, which is consistent with the lack of CTCF. The X chromosome contains weaker compartments along with loop-anchored TADs, formed by a condensin I variant that functions in dosage compensation (hereafter DC) in hermaphrodites (Crane et al. 2015, Kim Jimenez et al. 2022). At the multi-kilobase scale, we observe contact enrichment patterns that orthogonally protrude from the main diagonal across all chromosomes, which are termed "jets" or "fountains" and are formed by cohesin (Kim et al., Isiaka et al 2024). 

Condensin DC forms loop-anchored TADs on the X chromosomes 

C. elegans X chromosomes contain loop-anchored TADs, which are analogous to those formed by cohesin and CTCF in mammals. Through a series of deletion and ectopic insertion experiments in Kim Jimenez et al 2022, we proposed a model that condensin DC enters at the recruitment elements on the X (rex) sites and loop extrudes in one-direction. The rex sites also function as a block to condensin DC incoming from other sites (Figure 2). Next, we want to find out how rex sites function to both recruit and block condensin DC. It is possible that one or more of the recruiting proteins (SDCs and DPY-30) block condensin in a CTCF-like manner or the outgoing molecules' collision with the incoming ones prevent translocation across the loading sites. 

Cohesin loads and extrudes DNA loops originating near active enhancers and transcribed genes 

In recent work, we showed that cohesin forms Hi-C features termed "fountains", which are population average reflections of DNA loops originating from distinct genomic regions and are ~20-40 kb in C. elegans (Kim, Wang, Ercan 2024). Hi-C analysis upon cohesin and WAPL-1 depletion support the idea that cohesin forms these structures by preferentially loading and loop extruding in an effectively two-sided manner (Figure 3). The putative cohesin loading sites are active enhancers bound by the C. elegans ortholog of NIPBL and fountain patterns correlate with transcription and the distribution of NIPBL binding sites. 

Fountains are also observed in mouse and zebrafish (Galitsyana et al. 2024). We propose that preferential loading and loop extrusion by cohesin is an evolutionarily conserved mechanism that regulates the 3D contacts of enhancers. Compared to mammals, average processivity of C. elegans cohesin is ~10-fold shorter and the binding of NIPBL ortholog does not depend on cohesin. It is possible that CTCF and the longer processivity of mammalian cohesin have evolved with the increasing genome size and greater distances between enhancers and promoters. 

Although acute depletion of cohesin in L3 larvae did not show significant changes in mRNA-seq, Meister lab showed that the transcript isoform of skn-1 in head neurons change upon depletion of cohesin from L1 to L3 (Isiaka et al 2024). It is possible that cohesin plays a role in selection of alternative transcription start sites or functions earlier in development while differentiation processes set up epigenetic marks to control cell-type specific transcriptional programs. In current work we are addressing the function of cohesin in gene regulation. 

During cell division, as condensins leave the chromosomes during G1, cohesin loads to and organizes the interphase genome. In C. elegans, cohesin and condensin DC are both present on the interphase X chromosomes. We are curious how these two loop extruders that organize the 3D genome at two different scales (fountains and TADs) interact with each other. 

Mechanisms of SMC-mediated X chromosome dosage compensation in C. elegans 

In many animals, females have two and males have one X chromosome, while autosomes are present equally in both. Dosage compensation balances the expression of X chromosomes to that of autosomes, also equalizing X chromosomal gene dosage between XX females and XY males.  In humans, the Y chromosomes are small and carry few genes. In C. elegans Y chromosome is completely degenerated (shown as O). As in a few other nematode species, C. elegans XX animals are hermaphrodites that produce both sperm and oocytes. In humans, D. melanogaster, and C. elegans dosage compensation is essential and is established during embryogenesis. In C. elegans around the 40-cell stage, a hermaphrodite-specific protein, SDC-2 binds to the X chromosomes and recruits the rest of the dosage compensation complex (DCC) members in somatic cells (Dawes et al. 1999). The core of the DCC is condensin DC, formed by four condensin I subunits and an SMC4 paralog DPY-27 (Figure 4). In addition to SDC-2, DPY-27 interacts with SDC-1 and SDC-3 (two zinc finger proteins), DPY-30 (a protein that is also a subunit of the evolutionarily conserved MLL complex), and DPY-21 (a H4K20me2 demethylase). DCC binding represses transcription across the X chromosomes by two-fold in XX hermaphrodites equalizing to that of XO males. (For a more comprehensive review and citations see Albritton and Ercan 2018)

Chromosome-specific recruitment of an SMC complex 

In vitro purified condensins bind to DNA without aid (e.g. Terakawa et al). However, ChIP-seq analyses in multiple species show that condensins are not enriched uniformly across the genome (an early review: Jeppsson et al), thus there must be mechanisms that control their recruitment in vivo. This is particularly apparent for condensin DC, as it binds to and regulates the X chromosomes for dosage compensation. 

Experiments injecting different parts of the X chromosomes and analyzing their ability to recruit the DCC found that the complex is recruited at distinct recruitment elements on the X (rex) then spreads along the chromosome (Csankovszki et al. 2003, McDonel et al. 2006). In Ercan et al.  2007, we mapped the DCC binding sites and identified a DNA sequence motif enriched at the rex sites. In Albritton et al. 2017, we addressed why the motif, although less frequent, is on the autosomes but do not recruit the DCC there. We found that a single rex inserted on chromosome II recruited ~20% of what it does when on the X. Insertion of two other rexes ~30 and ~50 bk away from the single ectopic rex, increased its recruitment capacity to ~40%. Together with rex deletions, we concluded that long-range cooperation between rex sites robustly and specifically recruit the DCC to the X chromosomes. 

After sequence-specific recruitment, the DCC spreads linearly along the chromosome (Figure 5). The mechanism of linear spreading is consistent with loop extrusion and is observed as DCC spreading onto autosomal portion of the X-autosome fusion chromosomes by ChIP (Ercan et al. 2009), by mRNA-seq (Street et al. 2019) and by Hi-C (Kim Jimenez et al 2022).  We propose that cooperative recruitment is evolutionarily robust, because mutation of a motif on the X or presence of one on the autosomes do not disrupt the dosage compensation process, which is essential for embryogenesis. Similarly, spreading onto physically attached DNA may ensure dosage compensation of translocated autosomal DNA during genome evolution, as observed in our mRNA-seq analyses of different translocations in Ragipani et al 2022.  

Topoisomerase II is required for the in vivo processivity of condensin DC translocation 

X-autosome fusion experiments showed that condensin DC translocates linearly from the X into the fused autosome and the level of condensin DC reduces with distance from the fusion site (Ercan et al. 2009). Our analyses of condensin DC binding and activity using ChIP-seq and Hi-C upon TOP-1 and TOP-2 depletion revealed two distinct modes of condensin DC translocation (Morao et al. 2022). The first one is long-range, translocating over hundreds of genes, and is consistent with loop extrusion and requires the removal of DNA knots/catenates by Topoisomerase II (Figure 6). The second one is short-range, translocating across active gene bodies, and requires removal of positive supercoiling at the 3' of transcribed genes by Topoisomerase I. In follow up work, we are studying the role of both topoisomerases in promoting elongation by RNA Polymerase II.

A non-catalytic function for histone H4K20me2 demethylase DPY-21 in dosage compensation

C. elegans condensin DC interacts with DPY-21, a demethylase that removes a single methyl from H4K20me2, increasing H4K20me1 (Brejc et al. 2017). H4K20me1 is a highly conserved mitotic mark. In a null mutant of dpy-21, H4K20me1 level on the X reduced by 2-fold and RNA Pol II binding to X chromosomal promoters increased (Kramer et al. 2015). TADs and DNA loops on the X chromosomes did not differ significantly between the null mutant and wild type (Breimann, Morao et al. 2022). In collaboration with the Woehler and Preibisch labs at MDC Berlin, we performed fluorescence recovery after photobleaching (FRAP) experiments. Unlike the catalytically dead mutant, dpy-21 null mutant showed a significant reduction in the percentage of mobile condensin DC (Figure 7), thus a non catalytic activity of DPY-21 is required for condensin-DC mediated repression of the X chromosomes (Breimann, Morao et al. 2022). 

Our work is also an important example of a histone modifier participating in a transcriptional regulatory process without its catalytic function.

How does condensin DC repress transcription?

The mechanism by which condensin DC reduces RNA Polymerase II initation is still unknown. On X-autosome fusions, we observed that the level of condensin DC binding correlates with the level of repression (Street et al. 2019) . The dpy-21 null mutant derepresses the X chromosomes and reduces the proportion of mobile condensin DC molecules (Breimann, Morao et al. 2022). But we do not know if the mobile fraction  corresponds to those molecules loop extruding, translocating across genes, or is engaged in another type of behaviour. 

It is intriguing that condensin DC represses the entire chromosome and the level of reduction is ~2-fold across genes transcribed at differing levels (Kramer et al. 2016).  In Street et al. 2019, we found that condensin DC is enriched at gene regulatory elements and specifically regulates histone modifications associated with active transcription but not repressive chromatin. In Morao et al. 2022 we found that both RNA Pol II and condensin DC accumulate at the 3' end of genes upon TOP1 depletion. Strong correlation between active transcription and condensin DC binding suggest that they are coupled. 

To understand how condensin DC "fine tunes" transcription, we collaborated with Preibisch lab, and performed single molecule RNA FISH experiments in embryos (Breimann et al. 2024). Our results suggest that condensin DC reduces the frequency of transcription initiation (Figure 8). Current work in the lab addresses the mechanisms by which condensin DC binding or activity may be coupled to and in turn reduces the frequency of transcription initiation. 

Evolution of SMC-mediated X chromosome dosage compensation in two nematode lineages

Sex chromosomes evolved multiple times and so did a variety of X chromosome dosage compensation strategies (reviewed by Gu and Walters 2017). In the three well-studied model species with independent sex chromosome origins, three different mechanisms are employed. In Mus musculus (mouse) and humans, a long noncoding RNA, Xist acts as a scaffold for heterochromatin proteins and DNA methylation to inactivate one of the X chromosomes in XX females. In Drosophila melanogaster (fly), a histone acetylating ribonucleoprotein complex specifically binds to and activates transcription of X chromosomal genes by ~2-fold in XY males. In Caenorhabditis elegans (nematode), an X-specific Structural Maintenance of Chromosome (SMC) complex forms the dosage compensation complex (DCC) that represses transcription of X chromosomal genes by ~2-fold in XX hermaphrodites.

Despite our ability to sequence and use transcriptomics to assess dosage compensation in new species (e.g. Albritton et al. 2014), we are largely ignorant of the dosage compensation mechanisms in non-model species. As a result, we do not know how these mechanisms evolved after the initial descent of the X chromosomes and how they co-opted the same or different strategies of epigenetic regulation for dosage compensation. In Aharonoff et al. 2024, we started addressing these questions by anchoring our analyses to Caenorhabditis elegans, and exploring the surrounding phylogeny.  

Our results showed that the duplication and divergence of the SMC-4 paralog DPY-27 occurred suprisingly recently in the lineage leading to Caenorhabditis (Figure 9). Nematode species lacking two copies of the SMC-4 dosage compensate using other mechanisms. A second X-specific SMC complex evolved through an independent duplication within Pristhionchus. Current projects aim to determine the diversity of dosage compensation mechanisms in most lineages of nematodes. We are also curious to find if the X-specific SMC in P. pacificus co-opted the same or different epigenetic mechanisms for transcription regulation.

Other directions / projects in the lab

Environmental control of gene regulation: Environmental control of gene expression is key for species survival. C. elegans dauer larvae is a fascinating and metabolically different state that allows worms to withstand lack of food for months. We are using genomic techniques to address if dauers compact chromatin to achieve genome-wide gene repression and how this is linked to pathways that control dauer physiology.  

TFIIIC and ETCs in genome organization and gene regulation: TFIII is a general initiation factor for RNA Pol III, which transcribes tRNAs and other small RNAs. TFIIIC  binds to the evolutionarily conserved extra TFIIIC sites (ETCs) without promoting transcription. Previous studies proposed that ETCs function as a boundary between TADs or heterochromatin domains. We are testing this model by depleting TFIIIC and analyzing its effect on genome organization and gene expression. 

Mechanisms that regulate accurate chromosome segregation: In all organisms studied to date, a common phenotype of mutants that affect chromosome segregation are DAPI-staining DNA that stretches between segregating chromosome masses in anaphase. These “anaphase bridges” are poorly characterized, in part due to the assumption that their content is invariant or random. We are developing Bridge-seq to characterize anaphase bridges with the goal to connect them to genome integrity problems mapped in cancers and aging.