University of Edinburgh
Exploring mechanisms and impact of transcriptional compensation in craniofacial disease models of SOX9 haploinsufficiency
Abstract
dc:description.abstractVertebrate facial structures are derived largely from cranial neural crest cells (CNCCs) during development. These transient facial progenitor cells arise from the neural plate border region, travel long distances and differentiate into a wide array of derivatives including bone, cartilage, connective tissue, neurons, glia and pigment cells. Genetic and environmental perturbations of human CNCC development manifest in craniofacial malformations. In fact, one third of all developmental disorders have manifestation in the face. Haploinsufficiency of a key developmental transcription factor (TF), SOX9, causes campomelic dysplasia (CD), associated with bowing of the long bones, Pierre Robin sequence (small lower jaw, downward displacement of the tongue and upper airway obstruction), cleft palate, midfacial hypoplasia, and disorders of sex development. Interestingly, SOX9 plays a dual role in craniofacial development, firstly in the neural crest, but also later in development, as a master regulator of chondrogenesis. In this thesis, I aimed to uncover molecular mechanisms driving craniofacial disease phenotypes in CD using in vitro human cellular models of SOX9 haploinsufficiency. I therefore leveraged human embryonic stem cell (hESC) lines with heterozygous SOX9 mutations and employed a well-established in vitro protocol to differentiate them to CNCCs. Despite the engineered deletions causing an early frameshift and predicted loss of function, SOX9 protein levels in mutant CNCCs appeared equivalent to wild type cells, indicative of a buffering mechanism. Indeed, at the transcriptional level, I observed that expression of the unedited SOX9 allele in heterozygous mutant CNCCs was equivalent to SOX9 expression in wildtype cells. Based upon this surprising observation, I hypothesised that SOX9 gene expression is tightly controlled during CNCC differentiation such that haploinsufficiency is compensated by transcriptional upregulation, perhaps via a regulatory feedback loop. Initially, I explored the utility of mouse and human gastruloid protocols for modelling early neural crest development, which inspired a new Array-CNCC protocol based on an initial single-cell aggregation step. This adapted CNCC differentiation protocol, developed as part of this thesis in collaboration with Ewa Ozga, is intended to improve an existing in vitro differentiation protocol to standardise the early stages of CNCC emergence, improve reproducibility and facilitate co-culture of wildtype and mutant cells. Next, I sought to characterise the observed compensation of SOX9 expression at the RNA and protein level in heterozygous mutant CNCCs. I observed that by the late post-migratory CNCC stage, there were minimal transcriptional changes between wildtype and SOX9 mutant cells by RNA-seq, suggesting that the compensation had reached a steady-state level. I therefore decided to trace back to earlier stages to determine the timeframe for this upregulation. Through a combination of digital PCR (dPCR) and RNA-seq, I observed that transcriptional buffering first occurs during early stages of in vitro CNCCs induction as the cells migrate from neural spheres, reaching the peak by Day 11 of the differentiation, and then persisting through later development. Finally, I investigated possible mechanisms driving transcriptional compensation and TF dose-regulation in CNCCs. First, I determined that the observed upregulation of SOX9 mRNA levels in heterozygous cells is likely driven by increased transcription rather than increased stability of steady state mRNA. Furthermore, I found that inhibition of the nonsense mediated decay pathway (NMD) did not reverse SOX9 transcriptional compensation in heterozygous mutant cells, suggesting that this is not an NMD-dependent process such as transcriptional adaptation. Based on the small number of genes differentially regulated in my RNA-seq data, I took a candidate-based approach to test the possibility of a negative feedback loop involving SOX9 upregulating a transcriptional repressor that acts to suppress SOX9 expression. However, I found that perturbations of a putative repressor, ZNF558, did not impact SOX9 expression. From additional experiments in which I transiently overexpressed SOX9 in heterozygous mutant cells for a short time window, I observed downregulation of endogenous SOX9 expression, raising the possibility that SOX9 may directly autoregulate its own expression. Based on this, I proposed that SOX9 may compete with another activator for shared binding sites. These binding sites could therefore become occupied upon SOX9 down-regulation, leading to upregulation of SOX9 expression. A strong candidate for this from the literature is the LEF1/TCF/β-catenin complex which has been shown to have a shared DNA binding motif with SOX9, with one or the other able to bind to the site in a mutually exclusive manner. Additionally, SOX9 has been shown to inhibit canonical Wnt signalling pathway by direct binding to β-catenin. From preliminary experiments, I observed that inhibition of the Wnt pathway can reverse SOX9 transcriptional compensation in early migratory CNCCs, providing tentative insights into the mechanisms driving this phenomenon. Together, this thesis documents in detail the observation of autosomal gene dosage compensation in a hESC-based cellular model of disease for a key developmental transcription factor. Shedding light on this mechanism could explain how cells sense and buffer gene dosage in development and could be relevant to other transcription factors. Furthermore, I have explored cellular and molecular phenotypes during craniofacial development for human cellular models of campomelic dysplasia. Additionally, a new updated ARRAY-CNCC protocol holds promise to improve reproducibility and facilitate the study of early stages of human CNCC emergence in vitro. Understanding the sensitivity of the CNCC gene regulatory network to transcription factor expression levels and mechanisms of gene expression compensation in early development will be crucial for understanding disease aetiology and phenotypic variation for patients affected by mutations at the SOX9 locus.
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Milto, Katarzyna Maria
- Advisors dc:contributor.advisor
-
- Nichols, Jennifer
- Long, Hannah
Subjects
dc:subject × 8Rights
- Language dc:language.iso
- en
Identifiers
dc:identifier.*- Identifier URI
- https://doi.org/10.7488/era/7637