Sex-determining 3D regulatory hubs revealed by genome spatial auto-correlation analysis

Irene Mota-Gómez, Juan Antonio Rodríguez, Shannon Dupont, Oscar Lao, Johanna Jedamzick, Ralf Kuhn, Scott Lacadie, Sara Alexandra García-Moreno, Alicia Hurtado, Rafael D. Acemel, Blanche Capel, Marc A. Marti-Renom, Darío G. Lupiáñez

Posted on: 13 December 2022

Preprint posted on 18 November 2022

Like Meta, but better. METALoci: a new tool to identify 3D regulatory regions in the genome, sex-determination edition.

Selected by Martin Estermann

Categories: bioinformatics, cell biology, developmental biology, genomics, molecular biology

Background:

During early embryonic development, the gonads of both male and female embryos develop similarly: they contain the same cell types and show highly similar gene expression. During a process known as gonadal sex determination, the bipotential gonad differentiates into two morphologically different organs: ovaries in XX females and testes in XY males. This process is regulated by two mutually exclusive genetic programs. The testicular program is initiated by the upregulation of the Y-linked gene Sry in XY supporting cells, which activates Sox9 expression. Sox9 then activates Fgf9, which is important in maintaining Sox9 expression, generating a positive feedback loop and promoting the differentiation towards Sertoli cells. In XX females, the lack of Sry results in the upregulation of Foxl2, initiating the differentiation towards the pre-granulosa fate and inhibiting Sox9 and Fgf9 expression.

Mutations on these key genes can result in differences of sex development (DSDs). DSDs include a wide range of conditions associated with a discordance between the genetic and phenotypical sex. The first step in the genetic diagnosis of DSDs involves the evaluation of a set of pre-selected mutations that are traditionally associated with DSDs. If unsuccessful, next-generation sequencing (NGS) technologies are used to identify novel mutations. Unfortunately, around 50% of DSDs patients won’t receive a definitive diagnosis. One of the biggest caveats of using these NGS technologies is the large amount of information that they generate, making it difficult to efficiently detect the mutation that can explain the pathogenesis. Additionally, whole-exome instead of whole-genome sequencing is typically performed, focusing only on the coding region of the genome, and excluding the non-coding or regulatory regions from the analysis. New technologies are needed to filter the large amount of data generated by NGS methods and associate the data with a potential phenotype.

In this preprint, the authors developed a methodology to identify 3D enhancer regions during gonadal cell differentiation by combining Hi-C and ChIP-seq data. Additionally, they were able to model in silico how different mutations affect gene expression and validate these results using animal models.

Key findings

1) METALoci: a new tool to identify 3D enhancer regions in the genome.

By integrating Hi-C and epigenetic data (H3K27Ac), the authors developed a new tool to identify regions in the genome that are physically close, as result of the 3D organization of the chromatin, and regulatory active. The authors used this methodology to identify changes during testicular differentiation, from supporting cell progenitors (E10.5) (Figure 1A) to Sertoli cells (E13.5) (Figure 1B). As can be seen, despite the 3D chromatin structure not changing much, there is a differential accumulation of epigenetic marks associated with open chromatin and active gene expression in the Sox9 locus after sex differentiation. This is consistent with previous data showing that Sox9 is a key driver of testicular differentiation. Additionally, the two main enhancers known to regulate Sox9 expression (Enh13 and TESCO) were also located in close proximity to Sox9 and displayed active chromatin marks. This data demonstrates the power of this tool to identify enhancer regions in the 3D space, correlated with gene expression.

Figure 1. METALoci changes during Sertoli cell sex differentiation. Graphical representation in 2D of the organization of the chromatin before (A) or after (B) sex differentiation. Here, a region of chromosome 11 is shown, based on the Hi-C data (left) and including the H3K27Ac data (right). This chromosomal region contains the Sertoli cell marker Sox9 and its reported enhancers TESCO and Enh13.

2) Identification of novel regulatory regions in gonadal differentiation.

Another important gene in testicular differentiation is Fgf9. Before this study, it was unknown how this gene is genetically regulated. Using METALoci, the authors identified a potential regulatory region. They computationally eliminated different 50kb regions from the locus and predicted how these mutations would potentially affect its 3D regulation and thus Fgf9 expression. This allowed the identification of a potential regulatory non-coding region located 250kb downstream of Fgf9. To validate these results, the authors generated a mouse model with a deletion in this genomic region (D306) and evaluated the effects of this deletion on gonadal sex differentiation. As can be seen in Figure 2, the deletion of this region resulted in the development of ovotestis or ovary-like structures in XY (male) individuals, expressing both testicular (SOX9) and ovarian (FOXL2) gonadal markers, at different degrees. This phenotype was the same as the one described for Fgf9-null mutants, confirming the importance of this regulatory region in controlling Fgf9 expression.

Figure 2. Gonadal differentiation in wild type and mutant embryos. Immunofluorescence for testicular (SOX9) and ovarian (FOXL2) supporting cell markers in E14.5 gonads from wild type male (XY, left), female (XX, right) and XY embryos with a deletion in Fgf9 enhancer region (middle).

Why I chose this paper:

As someone who studies gonadal differentiation and is interested in the diagnosis of DSDs, this paper resonated with me. I think that the technology the authors developed here can be used to close the gap between the genotype and the phenotype, one of the biggest issues in DSD diagnosis. I would love to see what other novel regulatory regions are detected using this tool. Additionally, the methodology seems to be straightforward enough to be applied not only to the gonadal development, but to other organs or even organisms. I can’t wait to see new applications for this tool.

Future directions / questions for the authors:

What are the minimum requirements to perform this kind of analysis? Could you combine Hi-C with ChIP-seq of other epigenetic marks, like H3K27me3, H3K9Ac or H3K4me?

Do you think that the METALoci tool can be applied, in the future, to genome sequencing data from DSDs patients to identify what mutations are responsible for changes in gene expression and thus the development of DSDs? What do you think would be necessary to achieve this goal?

doi: https://doi.org/10.1242/prelights.33270

Read preprint