A molecular understanding of transposon-based enhancer activation by the ChAHP complex during human cell fate decisions

Transposable elements, or transposons, are short, repetitive sequences within a cell's DNA. They are sometimes known as "jumping genes" because they can replicate and integrate into new sites. These integrations are inherently mutagenic so our genomes have evolved mechanisms to repress them. This has resulted in an evolutionary arms race where transposons evolve to evade these defence mechanisms, while our defences evolve to silence the transposons. The result is that our genomes contain many thousands of inactivated, often mutated transposons, along with a relatively small number of active ones. One way a repeat might escape inactivation is to deliver useful function for the cell. Indeed, several examples exist where a recently evolved class of short interspersed nuclear element (SINE) is used by mammalian cells to enhance gene expression during development. These sequences have been termed "eSINEs".

Cells must precisely control when and where eSINEs are active lest uncontrolled activity derail normal development. The activity of DNA is controlled by how it is packaged within the nucleus. DNA is not "naked" within cells but exists in the form of chromatin, i.e., it is wrapped around proteins so it is compact and stable. A group of proteins called chromatin remodellers control how tightly packed different parts of the genome are, and therefore whether they are "active" or "inactive."

We have spent many years working on a chromatin remodeller called CHD4. CHD4 is a component of two different multi-subunit complexes, NuRD and ChAHP. We've shown that NuRD facilitates developmental decisions in mouse and human stem cells by controlling the activity of regulatory DNA sequences. We expect NuRD and ChAHP to share some similarities in how they act, however the two complexes have many different components and while NuRD is found at all active regulatory sequences, ChAHP localises mostly to SINEs.

Mutations in one of the ChAHP components, ADNP, give rise to a human neurodevelopmental disorder called Helsmoortel Van der Aa Syndrome (HVDAS). HVDAS is characterised by facial dysmorphisms, cardiovascular and gastrointestinal problems and autism spectrum disorder. Understanding what ChAHP does during human development and how it does it will provide important information not only for HVDAS, but also for understanding autism spectrum disorders and human tissue development generally.

We hypothesise that ChAHP holds eSINEs inactive but ready to be activated if/when cells need to start differentiating. We further suggest that ADNP is displaced from eSINEs during their activation, allowing CHD4 to recruit NuRD components which help the eSINEs to interact with promoters and activate gene expression.

In this project we will use cutting edge technologies to determine the molecular and developmental functions of ChAHP during human cell fate decisions. In mouse stem cells ChAHP acts on a class of SINEs which does not exist in primates so to fully understand human ChAHP function we need study it in human cells. We will create human pluripotent stem cells in which we can quickly deplete proteins and assess the primary consequences of their loss on ChAHP assembly and function. We will define how ChAHP is directed to its sites of action, what it does there, how it interacts with the cell's transcription machinery, and how ChAHP facilitates the use of eSINEs to control gene expression. We will define exactly which developmental decisions show ADNP dependency during formation of some of the tissues most affected in HVDAS, and then determine what goes wrong in cells harbouring disease-causing ADNP mutations. Together this will be a comprehensive investigation into how human stem cells are able to use eSINEs and how this can go wrong in human disease. Our findings will be of relevance not only to those affected by HVDAS or autism, but also to basic scientists studying transposons, transcriptional control, and human stem cell biology.

Changes to cell states which occur during development require precise changes in chromatin structure and gene expression, failure of which can interfere with developmental transitions. The nucleosome remodeller Chd4 plays an essential role in control of gene expression during cell fate transitions in early mammalian embryos and during ES cell differentiation. Chd4 is best known as a component of the Nucleosome Remodelling and Deacetylation (NuRD) complex, which acts to control chromatin structure and nuclear dynamics of enhancers to modulate active transcription. Chd4 also associates with the zinc finger/homeodomain protein ADNP and Heterochromatin protein 1 (HP1) in a different protein complex termed ChAHP. Mutations in ADNP give rise to the neurodevelopmental disorder Helsmoortel van der Aa Syndrome (HVDAS), and are one of the most frequent causes of familial autism. Defining ChAHP function will hence be important for better understanding both autism spectrum disorders and the developmental abnormalities seen in HVDAS. ChAHP associates with newly evolved short interspersed nuclear elements (SINEs) and is implicated in their use as enhancers. As recently evolved repetitive elements differ in humans and mice, mouse cells are of limited use for modelling human ChAHP function. Here we propose a comprehensive molecular, temporal and functional dissection of ChAHP function in human pluripotent stem cells. We will test our hypothesis that ChAHP represses, but licenses certain SINEs to be used as enhancers, and that upon activation these "eSINEs" recruit NuRD to facilitate their ability to enhance transcription. Using directed differentiation of pluripotent stem cells we will identify ADNP-dependent cell fate decisions potentially underlying developmental defects in HVDAS. This project will deliver a molecular and temporal understanding of ChAHP function and eSINE activation and identify developmental decisions impaired in human genetic disease.