The epigenetic regulation of cerebellar granule cell development

Student thesis: Doctoral ThesisDoctor of Philosophy


Cell-type specific transcription is controlled by dynamic interactions between gene promoters and long-range cis-regulatory elements, such as enhancers. In turn, the activity of cis-regulatory elements is modulated by the recruitment of lineage-specific transcription and chromatin remodelling factors. Cerebellar granule cells are the most abundant neurons in the central nervous system. The major steps and transcriptional changes that accompany the different stages of cerebellar granule cell precursor (GCp)-development have been relatively well-characterised. Yet, the cis-regulatory elements that coordinate the expression of genes with critical functions during GCp-development remain largely uncharted. 
To generate a genome-wide atlas of long-range promoter interactions at restriction fragment resolution in GCps, I performed promoter capture Hi-C involving 22,047 annotated gene promoters using freshly isolated proliferating mouse GCps. We identified 83,224 unique significant promoter contacting fragments in cis, of which ~22% contained histone modifications associated with enhancer activity. The target genes of some of the most significant promoter contacting fragments identified were both Zic1 and Zic4. These genes play an important role in coordinating the appropriate expansion of the GCp-pool; however, the cis-regulatory elements that coordinate their expression in GCps are largely unknown. Zic1 and Zic4 are hypothesised to be divergently transcribed from an intervening promoter and to share regulatory elements. In this thesis, I provided the first experimental evidence for common long-range enhancers of Zic1 and Zic4
The transcription factor MATH1 is as a master regulator of GCp-development. It is thought that MATH1 is recruited to cis-regulatory elements, however, the target genes of these elements are mostly not known. In this thesis, I demonstrated that MATH1 was recruited to and sufficient to induce activity of a Zic1/Zic4 enhancer(s) and thereby likely modulates transcription of these genes. Consistently, transcription of both Zic1 and Zic4 has been reported previously to be significantly downregulated in Math1-depleted GCps. I hypothesise that this interaction may be one of the mechanisms whereby MATH1 serves to promote the correct expansion of GCps during development. 
Both our group and others have previously established a critical role for the chromatin remodelling factor CHD7 in GCps. CHD7 is required for the appropriate expansion of GCps and to maintain chromatin accessibility at regions in the genome mostly distal to transcriptional start sites. The expression of this chromatin remodelling factor is upregulated in differentiating GCps, yet the role it plays during differentiation of this progenitor-pool remains unexplored. As a first step to determine the role of Chd7 during the differentiation of GCps, I deleted Chd7 at the point at which these cells initiated differentiation. I demonstrated that the previously reported increased expression of Chd7 in differentiating GCps is not reflected in an increased requirement of this chromatin remodelling factor at this stage for cerebellar growth. To the contrary, cerebellar hypoplasia in these mutant mice was distinctly milder compared to Chd7-deletion at earlier developmental time-points. A small number of reports have described that the initiation of differentiation and proliferation of GCps may not be mutually exclusive. Chd7-deletion at the point at which GCps initiated differentiation resulted in reduced proliferation of GCps at early postnatal stages. These data provided further evidence that the initiation of differentiation and ability to self-renew might not be mutually exclusive biological processes in GCps and that Chd7 in differentiating GCps may serve to maintain proliferation of the GCp-pool. Collectively, in this thesis I presented a genome-wide atlas of regulatory elements in GCps, functionally validated novel common cis-regulatory elements of Zic1 and Zic4, and identified a mechanism by which MATH1 likely modulates transcription of both these genes. Furthermore, I have shown that Chd7 in differentiating GCps is required for appropriate cerebellar growth, possibly by regulating proliferation of the GCp-pool.
Date of Award1 Feb 2020
Original languageEnglish
Awarding Institution
  • King's College London
SupervisorAlbert Basson (Supervisor) & Cameron Osborne (Supervisor)

Cite this