The Chromatin Remodeler and Tumor Suppress Chd5 Promotes Expression and Processing of Transcripts During Development of the Zebrafish Neural System
thesisposted on 14.05.2019, 17:41 by Erin L SorlienErin L Sorlien
Vertebrate neurogenesis is a multistep process that coordinates complex signaling pathways and chromatin-based regulatory machinery to generate highly specialized cells (Hsieh and Zhao 2016; Urban and Guillemot 2014; Alunni and Bally-Cuif 2016; Yao and Jin 2014; Schmidt, Strahle, and Scholpp 2013). Epigenetic factors play a fundamental role in underwriting neurogenesis in part by contributing to control of gene expression in differentiating neurons. A mechanistic understanding of the epigenetic machinery underlying neurogenesis in vertebrates is necessary both to fully understand biogenesis of neural tissue in this subphylum as well as to develop effective therapeutic strategies to treat diseased or damaged neural tissue.
An example of an epigenetic factor that is important for both neuronal differentiation and disease states is CHD5, a vertebrate-specific member of the CHD family of ATP-dependent chromatin remodeling proteins. Chromodomain / Helicase / DNA-binding (CHD) proteins play a variety of roles in vertebrate development, and misregulation or loss of CHD proteins has been linked to numerous diseases (Mayes et al. 2014; Marfella and Imbalzano 2007; Bartholomew 2014). CHD5 is expressed primarily in neural tissue, where it is thought to contribute to neurogenesis, and has been strongly linked to tumor suppression (Thompson et al. 2003; Vestin and Mills 2013). Loss of CHD5 plays a significant role in development of neuroblastoma, a devastating tumor that is a leading cause of cancer-related death in children (Jiang, Stanke, and Lahti 2011; Maris and Matthay 1999). Consistent with the disease phenotype associated with loss of CHD5, reduced expression of CHD5 impairs differentiation of neuronal cells (Egan et al. 2013b). However, ablation of CHD5 in mice surprisingly resulted in no detectable defects in brain development (Li et al. 2014; Zhuang et al. 2014). A subsequent report revealed that mice conditionally ablated for CHD5 in neural tissue exhibit symptoms consistent with an autism spectrum disorder (Pisansky et al. 2017). Much remains to be learned about the role of CHD5 in these processes to clarify these observations.
Further, Chd5 is unique among the family of Chd remodelers in that it provides a biochemical basis for crosstalk between the critical epigenetic marks H3K27me3 and DNA methylation. Chd5 and the closely related remodelers Chd3 and Chd4 are all components of the Mi-2/NuRD histone deacetylase complex that plays a critical role in mediating transcriptional repression in response to DNA methylation in mammals (Allen, Wade, and Kutateladze 2013). Only CHD5 is preferentially expressed in neural tissue, however, and only Chd5 remodelers have biochemical evidence of direct interaction with H3K27me3, which plays a critical role in enabling proper expression of transcriptional programs during neurogenesis (Egan et al. 2013b). Chd5 is thus unique among CHD remodelers in that it is biochemically linked to both DNA methylation and H3K27me3 in addition to being preferentially expressed in neural tissue.
With regards to mechanism, much remains to be learned regarding how Chd5 remodelers contribute to gene expression and tumor suppression. However, the data to date do not show extensive transcript phenotypes and it is not clear how the biochemical action of CHD5 contributes to the neurological phenotypes ascribed to altered expression of CHD5. Therefore, it is critical to determine a suitable context to study the role of CHD5 in these processes. Identification of CHD5-dependent genes in neurons is likely to generate insight into how loss of CHD5 contributes to tumorigenesis, in particular with regards to development of neuroblastoma. Regulatory pathways that drive neurogenesis have been found to be extensively conserved between humans and zebrafish. Therefore, we have turned to the power of the zebrafish model system to characterize how loss of Chd5 alters brain development during embryogenesis.
Importantly zebrafish development, and neurogenesis in particular, occurs largely over the first 5-days of development. Zebrafish are born outside of the mother, which can produce large clutches of several hundred embryos per week, providing us with an accessible context to study the role of chd5, the zebrafish homolog of human CHD5. The central nervous system of the zebrafish develops rapidly, and shares many of the organization features of the mammalian brain (Kalueff, Stewart, and Gerlai 2014). In particular, neuroblastoma arises from a population of cells known as sympathetic ganglion cells that are derived from the neural crest (Pei et al. 2013). These cells are conserved in vertebrates, and several models to study how these cells transform into neuroblastoma exist in zebrafish (Zhu et al. 2017; Morrison et al. 2016; Zhu and Thomas Look 2016). However, our understanding of the mechanisms controlling ganglion cell differentiation is incomplete and requires further investigation to understand how epigenetic and transcriptional mechanisms contribute to development of these cells and how failure of these processes leads to cancer. The neural crest undergoes a series of differentiation steps to form mature sympathetic neurons that are guided by bone morphogenic protein signaling, and transcription changes (Ernsberger and Rohrer 2018). These cells express key enzymes for synthesizing dopamine and norephinephrine to control the sympathetic system throughout the central nervous system (Ernsberger and Rohrer 2018).
To address these questions about Chd5, we have used CRISPR/Cas9 to generate chd5-/- zebrafish that are protein nulls as determined by western blot. These chd5-/- fish are phenotypically indistinguishable from wild-type fish under standard growth conditions as was previously observed for mice lacking CHD5 (Zhuang et al. 2014; Li et al. 2014). By using zebrafish, we are able to perform transcriptome analyses to identify Chd5 target genes at stages much earlier than has previously been performed in mice because we can harvest large amounts of the tissue of interest from the readily accessible embryos. We have therefore undertaken RNA-seq analysis of isolated brains from wild-type and chd5-/- fish to identify chd5-dependent genes in predominantly differentiating (2-day old) and substantially differentiated (5-day old) neural tissue. These data provide a substantively different perspective from previous studies that examine the role of CHD5 in gene expression of differentiated SY-SH5Y cells (Egan et al. 2013a) or in the forebrain of 8-week-old mice (Pisansky et al. 2017). (Jiang, Stanke, and Lahti 2011). One role we identified from this data, is the promotion of development of sympathetic ganglion cells (detailed below), illuminating for the first time a role for chd5 in promoting differentiation of cells directly involved in neuroblatoma.
We observe not only extensive changes in gene expression, but also identify a novel role for Chd5 in enabling proper splicing during this critical window of neurogenesis in the zebrafish brain. We are further exploring the role of CHD5 in these processes by creating comparable cell culture-based models of loss of CHD5 to determine the conservation of molecular phenotypes observed in zebrafish. Furthermore, this model enables us to leverage the extensive biochemical tools available in cell culture to examine alterations to the chromatin that are difficult to interpret from studies of complex tissues such as the brain.
Herein I will describe the research progress we have made to understand the role of Chd5 in gene expression and splicing in zebrafish, as well as ongoing work to engineer mouse embryonic stem cells as an additional model to study the transcriptional consequences of loss of CHD5. Critically, the addition of the cell culture model will greatly enable biochemical characterization of the changes that are leading in particular to the changes in gene expression and splicing and will provide us with a context to test for a direct role of CHD5 in these processes. In addition, this thesis will detail the results from ongoing projects using the zebrafish model system, including: development of models in zebrafish to study the tumor suppressive role of Chd5, phenotypes observed using a targeted chemical-genetic screen, and advancement in developing new tools in zebrafish to engineer specific genomic modifications that will greatly expand the power of this vertebrate model.