Research 2011-2020

Click here for: Lay readers' summary of our research

Compaction of eukaryotic chromosomes is dynamically regulated across the cell cycle. During mitosis, chromosomes are highly compacted to allow for accurate segregation. During interphase, chromosomes are decompacted but packaged to fit in the nucleus in an organized manner. Eukaryotic chromosomes are long linear fibers of DNA wound around the histone proteins forming nucleosomes like beads on a string. Chromatin fibers further fold and sit into individual chromosome territories within the nucleus. DNA packaging at each level regulates transcription. For instance, nucleosomes reduce DNA accessibility to transcription factors and RNA Polymerase. Looping of chromatin fibers mediate regulatory interactions between enhancers and promoters. How higher level of chromatin folding regulates gene expression is unclear. The goal of our lab is to uncover the molecular mechanisms that regulate chromosome structure and gene expression. The questions that drive our current research are: How are chromosomal domains specified and targeted for gene regulation? What are the transcriptional mechanisms that regulate multiple genes across a domain? What are the functional and evolutionary consequences of domain-scale gene regulation?

Majority of our work on chromosome structure and function is on X chromosome dosage compensation in Caenorhabditis elegans. Dosage compensation is an essential developmental process that equalizes X chromosome transcription between sexes in many animals, including humans. Although strategies differ between species, in the model organisms M. musculus (mouse), D. melanogaster (fly) and C. elegans (worm), multi-subunit dosage compensation complexes (DCC) bind to and regulate X chromosome transcription specifically in one sex. The DCC subunit compositions indicate that the different dosage compensation systems co-opted and targeted distinct sets of transcripton regulatory chromatin machineries to the X chromosome. The diverse strategies of X chromosome dosage compensation provide a unique and experimentally tractable window into the mechanisms that target and regulate transcription across large chromosomal domains.

In C. elegans, DCC specifically binds to and reduces transcription from both X chromosomes by half in XX hermaphrodites (Figure 1). The core of the DCC is a condensin complex, which belongs to the evolutionarily conserved structural maintenance of chromosomes (SMC) family of complexes. SMC complexes have a remarkable ring structure that form DNA loops. Defects in SMCs disrupt chromosomal functions in a wide range of developmental diseases and cancer.  Condensins are essential for chromosome compaction and segregation during cell division, and regulate genome organization and gene expression during interphase. Our research contributes to understanding how condensins bind to and compact DNA and how condensin-mediated DNA compaction affects transcription. Below is a summary of the current projects in the lab.

1. SPECIFYING AND TARGETING CHROMOSOMAL DOMAINS FOR GENE REGULATION

Coordinated regulation of genes across a chromosomal domain requires specific targeting of the regulatory machinery. Our findings on DCC binding, and research in heterochromatin silencing, Polycomb mediated gene regulation, and X chromosome dosage compensation suggest a unified strategy for targeting large chromosomal domains: recruitment and spreading. We address the mechanisms of DCC recruitment and spreading on the X.

1A. Specificity of DCC recruitment to the X chromosome

The core of the DCC is differs from the canonical condensin I by a single SMC4 variant (Csankovszki et al. 2009). It is remarkable that exchange of one subunit confers a new function to condensin. We call this complex condensin DC, for its role in dosage compensation (Figure 2). 

In C. elegans, condensins I and II bind to all chromosomes, but condensin DC is specifically targeted to the X. Condensin DC is recruited to the X in part by a short DNA sequence motif (GCGCAGGG) that is enriched at a small number of sites called the recruitment elements on the X (rex) (McDonal et al. 2006). Condensin II also binds to the rex sites, and this binding requires the DCC subunit SDC-2. Autosomal condensin II contains a partial motif of GCGC. We hypothesize that canonical condensins are recruited by both X and autosomal recruiters, but condensin DC can only be recruited by the X chromosomal recruiters comprising SDC-2, SDC-3 and DPY-30.

Interestingly, the 12-bp DNA sequence motif is important for X-specific DCC recruitment but the motif is found throughout the autosomes, but do not recruit the DCC there. Such context-dependent binding is not a problem unique to the DCC. Most transcription factors bind to a small fraction of their respective motifs. We found that insertion of a single recruitment site on an autosome is not enough to recruit the DCC there (Albritton et al 2017). In addition, deletion of a single recruitment site reduces DCC binding by ~20-40% (Figure 3). Our model is that the recruitment sites on the X cooperate to establish a robust domain of binding. Such cooperative action can establish X-specific binding without the need for establishing and maintaining specific sequences on the X.

1B. Mechanism of DCC spreading along the chromosome

After recruitment to the rex sites, DCC spreads along the the chromatin (Ercan et al. 2009). Unlike recruitment, spreading is not dictated by a DNA sequence motif, and can occur on autosomal DNA that is physically attached to the X. In strains that contain X;A fusion chromosomes, condensin DC spreads onto the autosomal region as far as 1-3 Mb (Figure 4). The spreading occurs mostly at accessible chromatin mapped by ATAC-seq (Street et al. 2018). Chromosomal binding characteristics of condensin DC and canonical condensins are similar, i.e. binding at distinct intergenic sites including promoters, enhancers and tRNAs (Kranz et al. 2013). The mechanism of spreading is unknown, and we are currently using the X;A fusion chromosomes and gene-editing techniques to determine how the DCC spreads linearly along the chromosome. 

2. TRANSCRIPTIONAL REGULATION OF MULTIPLE GENES WITHIN A DOMAIN

Expression of an individual gene is controlled by specific transcription factors that bind to the gene's regulatory regions. Instead of evolving a common regulatory element for each gene, domain-wide gene regulation controls the chromatin structure across multiple genes. In X inactivation, Xist spreading leads to enrichment of  heterochromatin marks that silence transcription. In C. elegans, DCC reduces RNA Pol II binding to promoters across the X. 

2A. Regulation of chromatin structure and transcription by the DCC

In collaboration with the Strome and Ahringer labs, we found that the DCC increases the level of H4K20me1 on the X by suppressing its conversion to H4K20me2/3 (Vielle et al. 2012). To understand the role of H4K20 methylation in DCC function, we studied chromatin structure and gene expression in H4K20 methyltransferase mutants (Kramer et al. 2015). The pattern of H4K20me1 ChIP-seq enrichment suggest that this modification is increased uniformly across the X . Since standard ChIP-seq and RNA-seq cannot analyze uniform increase or decrease across the genome, we developed and applied a spike-in normalization technique using C. briggsae as a standard. Our results showed that indeed H4K20me1 represses X transcription. However, two lines of evidence suggest that this might be indirect . First, DCC started repressing transcription in early embryos before H4K20me1 enrichment on the X. Second, the level of H4K20me1 was more important than its balance between X and autosomes. Our current hypothesis is that H4K20me1 regulates DCC binding to chromatin (Figure 5).Interestingly, condensin binding and H4K20me1 increases on mitotic chromosomes in all metazoans. Our work on the DCC and H4K20me1 could elucidate the link between this mitosis-enriched histone modification and condensin function.     

2B. Condensin-mediated repression of transcription

Since condensin DC binds to active promoters and enhancers, we wanted to know if this binding reduces the activity of these elements thus RNA Pol II initiation. In fact, knockdown of DCC resulted in increased H3K4me3 and H3K27ac, marks of promoter and enhancer activity (Street et al. 2018) (Figure 6). Interestingly DCC was not required not was affected by represisve marks, suggesting that the activity of the complex is tuned to transcription itself. This makes sense, as we had observed that the DCC binding correlates with transcription dynamically during development (Ercan et al. 2009). While active histone marks are associated with transcription, it is difficult to discern whether they are the cause or consequence of transcription activation itself. Next we will address this question using inducible protein degregation system that was recently applied to C. elegans.

2C. Condensin-mediated regulation of chromosomal interactions in yeast

Unlike metazoans, yeast contain only a single condensin. In collaboration with the Hochwagen lab, we performed genome-wide chromosome conformation capture and found that the yeast condensin mediates short-range interactions globally and long-range interactions at specific loci (Figure 7).  Interestingly, yeast condensin is required for interchromosomal interactions between tRNA genes, suggesting that in addition to the loop extrusion process that is likely mediating intrachromosomal short-range interactions, condensin may possess other mechanisms that mediate long-range interactions. 

3. EVOLUTION AND FUNCTION OF X CHROMOSOME DOSAGE COMPENSATION

In mammals, worms and flies, mutants deficient for dosage compensation invariably die during development, but it remains unclear why this is the case. To understand the essential function(s) of dosage compensation, we need to evaluate why dosage matters. For example, if all genes on the X chromosome tolerated two-fold higher expression in females, there would be no need to dosage compensate. We address the function of dosage compensation by studying the link between dosage compensation, sex-biased gene expression and X chromosome gene content.

3A. Dosage compensation and the evolution of X chromosome gene content

Males contain a single X chromosome and two copies of each autosome (XY, AA), while females contain a full set of chromosomes (XX, AA). Susumu Ohno hypothesized that the potential haploinsufficiencies unveiled by X monosomy in males were counteracted by increased expression from the single X chromosome. Ohno’s hypothesis predicts a similar transcriptional output from the single copy X and two-copy autosomes in males. Within a collaborative project, we verified that the average X expression is comparable to that of autosomes in mouse, fly and worm males (Deng et al. 2011). However, X to autosome comparison is complicated, because they harbor different genes. To address this, we mapped X chromosome in multiple nematode species (Figure 8), and compared the expression of ~300 one-to-one orthologs that are differentially located on the X and autosomes in C. elegans versus P. pacificus (Albritton et al. 2014). Our results suggest that not all X chromosomal genes were upregulated, as the orthologs located on the X were expressed less than their autosomal counterparts. Furthermore, predicted dosage sensitive genes are depleted from the nematode X chromosomes. This suggests that the monosomy of the X was counteracted by multiple mechanisms including upregulation of some dosage sensitive genes or their movement to autosomes.

3B. X chromosome dosage compensation and sex-biased gene expression

Genome-wide studies in various organisms including C. elegans indicate that the X chromosome harbors more genes with sex-biased expression compared to autosomes. However, these studies compound the effect of X chromosome copy number (dose) and sex by comparing the two sexes. For instance, by virtue of being in two copies in females and in one copy in males, X chromosomal genes can have 2-fold female-biased expression. We used well-characterized sex-reversal mutants in C. elegans to untangled the effect of X dose and sex (Figure 9) (Kramer et al. 2016).  Briefly, our results show that the X chromosome copy number increases hermaphrodite-biased expression of genes on the X in early embryogenesis and in the germline where several DCC subunits are not expressed. Hermaphrodite-biased expression due to X dose may be important for gonadal functions in adults. Therefore, we hypothesize that a conserved lack of dosage compensation in female gonads contributes to female-biased expression of X chromosomal genes with important reproductive functions. As a result, reproductive sterility in XO Turner syndrome patients may be due insufficient expression of female-biased genes on the X.