The evolution of the genome causes changes in gene expression which controls development and the regulation of these changes results in evolutionary modifications that consequently alter an organism’s function and adaptation. Studies of embryonic development in the sea urchin, Xenopus laevis, Caenorhabditis elegans, and other organisms have shown that developmental processes are a cascade of regulatory networks that determines developmental functions and these linkages have been conserved throughout evolution. While development has been extensively studied in the nematode C. elegans, less is known about the genomic architecture of development in other nematode species. We are interested in using C. elegans and distantly related entomopathogenic nematodes (EPNs), which are worms that parasitize and efficiently kill insects, from the genus Steinernema to study the similarities and differences in gene expression changes that give rise to similar body structures in these two species. First, we improved the genome and gene annotations of Steinernema carpocapsae to reliably conduct functional genomic analyses and to study responses to environmental changes via gene expression. Then, we compared two developmental stages, adults and infective juveniles (IJs), and the sexes of S. carpocapsae to equivalents in C. elegans. This comparison provided a set of conserved genes found in young adults and in another set in IJs. The comparison also gave insights into evolutionary changes to the regulation of gene expression which leads to similar morphological features. Subsequently, we profiled the transcriptomes of embryonic single cells from S. carpocapsae to identify founder cells that give rise to all the tissues in an adult worm. We used known C. elegans genes that determine the six founder cells, which give rise to all the body tissues, to identify these cells in S. carpocapsae and distinguish early embryonic cell fate. Lastly, to understand gene regulation we took a step towards C. elegans population genetics to emphasize within species variation in gene regulation. We compared the mRNA and microRNA profiles of twelve strains at the L1 stage. The comparison aimed to understand the differences in adaptation within one species to several differing environments. We found a set of 37 miRNAs that regulates gene expression at the L1 stage. Overall, all the approaches provide insights into the fundamental rules of gene expression that control changes at the genome and transcriptome level in development.

The complex task of maintaining homeostasis and fighting diseases involves an intricate network of immune cells with many relevant players. This thesis is focused on the plasticity and versatility of a critical class of innate immune cells called macrophages. Most naïve macrophages, named M0s, have the ability to polarize into two main subtypes, M1s and M2s, which help maintain a balance of inflammatory and anti-inflammatory responses, respectively. An imbalance in the ratio of M1s to M2s is associated with poor prognoses for a variety of diseases. Thus, understanding the markers and the gene regulatory networks (GRNs) that underlie the M0 to M1 or M2 polarization is crucial to help modulate these cells ratios for therapeutic purposes. Here, we applied bulk and single-cell RNA-seq and ATAC-seq to a high-resolution time series of HL-60-derived M0s polarizing towards M1 or M2 over 24 hours. We identified transient M1 and M2 markers and the main transcription factors (TFs) that drive polarization. In addition, we identified a novel M2 marker, ID2. We built bulk and single-cell polarization GRNs and identified at least 30 novel TF-TF interactions during M1/M2 polarization. We further compared the strengths of using bulk and single-cell technologies to build our GRNs providing experimental and computational guidelines for building GRNs of cellular maturation in response to microenvironmental cues. We concluded that despite the great advances of single-cell analysis, a combination of bulk and single-cell techniques provided a more complete GRN. The brain resident macrophages, named microglia, do not fit into the dichotomic M1/M2 dogma of polarization. However, microglial activation and inflammation are directly linked to progression of Alzheimer’s disease (AD). Neuroinflammation, hyperphosphorylated tau, and accumulation of amyloid beta plaques in the brain are hallmarks of AD, which presents progressive dementia as its main clinical feature. Amyloid plaques can activate the complement system. Complement activation, specifically activation of complement factor C5a and its receptor C5aR1 enhances microglial inflammation, which can worsen disease pathology through local injury and neuronal death. Thus, the C5a-C5aR1 signaling pathway is a potential target for modulation of AD. In order to investigate the effects of C5a in AD progression, we observed changes in hippocampal gene expression, hippocampal-dependent memory decline, and neuronal loss in two variants of the Artic mouse model of AD: one which lacks C5aR1 (cohort ArcticC5ar1KO) and one that overexpresses C5a under the GFAP promoter (cohort ArcticC5a+). The ArcticC5aR1KO group showed decreased inflammation, reduced activity of phagocytic and lysosomal pathways, and reduced cholesterol biosynthesis compared to Arctic mice. Furthermore, C5a overexpression led to poor cognitive performance, neuronal loss, and advanced disease progression compared to control. Our results suggest that pharmacological inhibition of C5a-C5aR1 signaling is a promising therapeutic strategy to treat AD.

Gene expression at the transcriptional level is controlled by DNA sequences called cis-regulatory modules (CRM) and at the post-transcriptional level by microRNAs (miRNAs). CRMs have been studied almost exclusively in bilaterian organisms and little is known about them in non-bilaterian metazoans. Understanding the architecture of CRMs in cnidarians, a sister phylum to bilaterians, can potentially shed light on the evolution of gene regulation. Head regeneration is one of the most widely studied developmental processes in cnidarians. Using a comparison of the transcriptomes of regenerating heads and developing buds, I have determined sets of genes that are specific and common between head regeneration and budding. To understand the genomic sequences controlling these developmental programs, I have mapped the open-chromatin landscape of Hydra in different body parts and during head regeneration to identify candidate promoters and enhancers. My results are the first atlas of CRMs in Hydra, including a substantial fraction that is dynamic during head regeneration.

Mammalian embryonic development has been used as a model system to study the role of miRNAs in previous studies, but a complete atlas of miRNA expression during development is missing. To understand the role of microRNAs during mouse development, I analyzed a time course of development representing multiple tissues and organs in mouse embryo. We find distinct tissue and developmental stage-specific miRNA expression profiles dominated by a small number of miRNAs. Analysis of conserved miRNAs reveals clustering of expression patterns by tissue types rather than species. We used matching RNA-seq and histone modification ChIP-seq datasets to improve the annotation of miRNA primary transcripts. We show that the expression levels of majority of primary miRNA transcripts predict the expression of their corresponding mature miRNAs. Our data provide the most comprehensive miRNA resource for mouse as well as a comprehensive list of mouse miRNAs that can be reliably measured by RNA-seq of their primary transcripts.

Taken together, the elucidation of cis-regulatory landscape in the cnidarian Hydra and miRNA expression during mouse embryonic development will help the scientific community to understand better the role of enhancers in metazoan evolution and miRNA regulation in mammalian embryonic development.

Facioscapulohumeral muscular dystrophy (FSHD) is a rare disease with characteristic weakness in facial and periscapular muscles which progresses to additional muscle groups. FSHD is caused by the misexpression of the embryonic transcription factor DUX4 in muscle cells. In 95% of FSHD patients, a series of macrorepeats preceding DUX4 is contracted and derepressed in part through loss of DNA methylation. In the remaining FSHD patients, the repeats are derepressed but not contracted, and 80% of these patients have a mutation in SMCHD1, which regulates DNA methylation in these repeats. DUX4 is involved in zygotic genome activation (ZGA) when it activates a number of transcription factors and chromatin remodelers, such as DUXA, LEUTX and ZSCAN4, as well as long terminal repeats (LTRs), such as ERVL-MaLRs, which are also activated in FSHD. DUX4 expression in patient muscle cells is sparse (0.5% of myotube nuclei), but its expression in only a few nuclei is sufficient to activate target gene expression in multiple nuclei within a multinucleated muscle cell, which is sustained when DUX4 is no longer present. My work has focused on understanding progression of FSHD at a molecular level both into different muscle groups and following DUX4 activation.I used single nucleus RNA sequencing to understand the contribution of individual nuclei to gene dysregulation following DUX4 expression. I identified nuclei with native expression of DUX4, as well as two populations of nuclei with high and low expression of DUX4-induced genes. The high group appears to perpetuate pathogenesis and has higher expression of genes related to the cell cycle despite the nuclei coming from cells in G0. I also found that DUX4 is coexpressed with only a subset of its target genes, while the DUX4 homolog DUXA is expressed with a wider set of targets. To understand why certain muscle groups are commonly or less affected in FSHD, I assayed DNA methylation and gene expression in different muscle groups. Genes induced during myogenesis in FSHD have higher expression in commonly affected muscle groups despite their promoters having high DNA methylation. Muscle groups differ in expression and DNA methylation of transcription factors key to developmental patterning and specification that may contribute to susceptibility to FSHD. Finally, I explored the role of DUXA as a potential regulator of DUX4 target genes following their initial activation. I found that DUXA depletion is sufficient to lower expression of DUX4 target genes including LTRs. I also identified a set of genes which are induced along with DUX4 during myogenesis in FSHD2 that are not induced following DUXA depletion. I have thus identified a candidate regulator of FSHD gene dysregulation and candidate contributors to differential muscle group susceptibility in FSHD.

Individual development is a complex process with a myriad of developmental controls at multiple levels ranging from individual cells to organs and entire individuals. The development and specification of each cell ultimately encoded in the genome. But whereas the genome is the same for all cells of the same individual, cell differentiation, specialization and response to the environment is regulated at the epigenetic level by gene regulatory networks (GRNs). Functional genomics studies have revealed that protein- DNA interactions, DNA methylation and changes in chromatin accessibility are essential to maintain cell identity and that interruption of these GRNs causes defects in cell development that can lead to disease and abnormal behaviors in individuals. Given the importance of epigenetic regulation in cells, tissues, and individuals, it would be interesting to know how these GRNs are conserved and evolve during mammalian evolution and how they can go wrong in disease. In the thesis, I present functional genomics studies and expand the understanding of epigenetic control in development from four aspects: (1) changes in DNA methylation in the same individual can be used as signature of different life experiences; (2) mutations in a repressor can cause abnormal gene expression in a small group of cells that further induce the onset of muscle wasting disease FSHD; (3) comparative dynamics of chromatin accessibility during definitive endoderm differentiation can identify conserved regulatory modules as well as species-specific enhancements; (4) The canonical form of the transcription factor NRSF is stabilized in genome through motifs conversion during mammalian evolution. These results show the versatility of epigenetic control during development and disease as well as highlight evolutionary forces shaping GRNs.

Alternative transcripts arise from the same gene via alternative TSS usage, splicing, and polyA site choice. Such transcripts can give rise to functional disparities in protein structure, post-transcriptional regulation, and translational efficiency. Moreover, their expression in appropriate spatiotemporal contexts is a key feature of eukaryotic genomes. However, detecting and quantifying these transcript isoforms across tissues, cell types, and species has been challenging due to their longer lengths compared to the short reads typical of standard RNA-seq. In contrast, long-read RNA-seq (LR-RNA-seq) provides complete transcript structures, enabling investigation of transcript features and usage with greater fidelity. Here, I describe my work on application of LR-RNA-seq to characterizing and comparing full-length transcriptomes. First, I describe Swan, a software library I developed to facilitate visualization of full-length transcripts and to compare transcript usage between biological conditions. Next, I describe the ENCODE4 human and mouse LR-RNA-seq datasets, where I applied a novel triplet-based framework to harmonize and classify transcripts that share transcript start sites, exon junction chains, and transcript end sites. Lastly, I discuss the application of our single-nucleus LR-RNA-seq technique (LR-Split-seq) on two geneticallydistinct mouse strains to uncover cell type and genotype-specific transcript usage patterns. Collectively, these projects form a solid foundation for future analyses of long read transcriptomes to quantify changes in transcript diversity and transcript usage between samples, cell types, and genotypes within and between species.