Plant embryogenesis is an intriguing developmental process that is controlled by many genes. AGAMOUS Like 15 (AGL15) is a MADS-domain transcriptional regulator that accumulates preferentially during this stage. However, at the onset of this work it was unknown which genes are regulated by AGL15 or how AGL15 is regulated. This dissertation is part of the ongoing effort to understand the biological roles of AGL15. To decipher how AGL15 functions during plant development, a chromatin immunoprecipitation (ChIP) approach was adapted to obtain DNA fragments that are directly bound by AGL15 in vivo. Putative AGL15 targets were isolated, and binding and regulation was confirmed for one such target gene, ABF3. In addition, microarray experiments were performed to globally assess genes that are differentially expressed between wild type and agl15 young seeds. Among them, a gene, At5g23405, encoding an HMGB domain protein was identified and its response to AGL15 was confirmed. Preliminary results suggest that the loss-of-function of At5g23405 might have an effect on somatic embryogenesis, consistent with AGL15 repression of the expression of this gene. Lastly, to address the question about how the regulator is regulated, the cis elements controlling the expression of AGL15 must be identified. Deletion analysis of the AGL15 promoter indicated the presence of putative positive and negative cis elements contributing to the expression of AGL15. Further analysis suggested that AGL15 regulates the expression of its own gene and this regulation may partially be explained by the direct binding of the protein to the AGL15 promoter. The data presented in this dissertation demonstrate that ChIP can be used to identify previously unsuspected targets of AGL15. Based on ChIP, a ChIP-chip technique is being developed in the lab to allow a more global analysis of in vivo binding sites. The identification of target genes and cis elements in AGL15 promoter is a step towards characterization of the biological roles of AGL15.

In eukaryotes, DNA is wrapped around a group of proteins termed histones that are required to precisely control gene expression during development. The amino acids of both the globular domains and unstructured tails of these histones can be modified by chemical moieties, such as methylation, acetylation and ubiquitination. The ‘histone code’ hypothesis proposes that specific combinations of these and other histone modifications contain transcriptional information, which guides the cell machinery to activate or repress gene expression in individual cell types. Chromatin immunoprecipitation (ChIP) experiments using undifferentiated stem cell populations have identified the genomic co-localisation of histone modifications reported to have opposing effects on transcription, which is known as bivalency. The human α-globin promoter, a well-established model for the study of transcriptional regulation, is bivalent in embryonic stem (ES) cells and this bivalency is resolved once the ES cells terminally differentiate (i.e. only activating or repressing marks remain). In a humanised mouse model, the deletion of a bone fide enhancer within the human α-globin locus results in heterogeneous expression patterns in primary erythroid cells. Notably, this correlates with an unresolved bivalent state at this promoter in terminally differentiated cells. Using this mouse model it is not feasible to ascertain whether the transcriptional heterogeneity observed in the cells lacking an α-globin enhancer is reflective of epigenetic heterogeneity (i.e. a mixed population of cells) rather than co-localisation of bivalent histone modifications within the same cells. Furthermore, the functional contribution of bivalency to development has yet to be described. To address these difficulties, I aimed to generate a fluorescent reporter system for human α-globin to facilitate the separation of transcriptionally heterogeneous erythroid cells. This model will provide material for ChIP studies on transcriptionally active and inactive populations to determine whether the epigenetic bivalency is reflective of a mixed cell population or true bivalency. In addition, I aimed to produce epigenetic editing tools to target bivalent promoters, which in combination with in vitro differentiation assays would provide an interesting framework to test the function of bivalency during development. In this study, I extensively tested gene-editing strategies for generating a fluorescent reporter knock-in in humanised mouse ES cells. I validated the suitability of humanised mouse ES cell lines for gene targeting studies and optimised a robust in vitro differentiation protocol for studying erythropoiesis. I utilised both recombineering and CRISPR/Cas9 gene editing tools in tandem with PiggyBac transposon technology, to knock-in the reporter gene. I made significant steps in gene targeting and successfully inserted the reporter downstream of the α-globin gene. I also generated a cloning system to express site-specific DNA-binding domains (TALEs) fused to epigenetic regulators with the aim to resolve bivalent histone modifications in vitro. From preliminary tests using these fusion proteins targeting Nrp1, a bivalent promoter in mES cells, I observed mild but significant changes in gene expression although histone modifications were unchanged. The various tools generated and tested in this study provide a solid foundation for future development of genetic and epigenetic editing at the human α-globin and other bivalent loci.

The main focus of this research was the development of microfluidic technology for ultrasensitive and fast molecular analysis of cells. Chromatin immunoprecipitation (ChIP) assay followed by next generation sequencing serves as the primary technique to characterize the genomic locations associated with histone modifications. However, conventional ChIP-seq assay requires large numbers of cells. We demonstrate a novel microfluidics-based ChIP-seq assay which dramatically reduced the required cell number. Coupled with next generation sequencing, the assay permitted the analysis of histone modifications at the whole genome from as few as ~100 cells. Using the same device, we demonstrated that MeDIP-seq with tiny amount of DNA (<5ng) generated high quality genome-wide profiles of DNA methylation. Off-chip sonication often leads to sample loss due to multiple tube transferring. In addition, conventional sonicators are not able to manipulate samples with small volume. We developed a novel microfluidic sonicator, which is able to achieve on-chip DNA/chromatin shearing into ideal fragment size (100~600bp) for both chromatin immunoprecipitation (ChIP) and methylated DNA immunoprecipitation (MeDIP). The integrated on-chip sonication followed by immunoprecipitation (IP) reaction can significantly reduce sample loss and contamination. Simple and accessible detection methods that can rapidly screen a large cell population with single cell resolution have been seriously lacking. We demonstrate a simple protocol for detecting translocation of native proteins using a common flow cytometer which detects fluorescence intensity without imaging. Using our approach, we successfully detected the translocation of native NF-kappa B (an important transcription factor) at its native expression level and examine the temporal dynamics in the process. Droplets with encapsulated beads and cells have been increasingly used for studying molecular and cellular biology. However, a mixed population of droplets with an uneven number or type of encapsulated particles is resulted and used for screening. We developed a fluorescence-activated microfluidic droplet sorter that integrated a simple deflection mechanism. By passing droplets through a narrow interrogation channel, the encapsulated particles were detected individually. The microcontroller conducted the computation to determine the number and type of encapsulated particles in each droplet and made the sorting decision. Our results showed high efficiency and accuracy for sorting and enrichment. / Ph. D.

Epigenetic modifications, such as DNA methylation and histone modifications, play important roles in gene expression and regulation, and are highly involved in cellular processes such as stem cell pluripotency/differentiation and tumorigenesis. Chromatin immunoprecipitation (ChIP) is the technique of choice for examining in vivo DNA-protein interactions and has been a great tool for studying epigenetic mechanisms. However, conventional ChIP assays require millions of cells for tests and are not practical for examination of samples from lab animals and patients. Automated microfluidic chips offer the advantage to handle small sample sizes and facilitate rapid reaction. They also eliminate cumbersome manual handling. In this report, I will talk about three different projects that utilized microfluidic immunoprecipitation followed by next genereation sequencing technologies to enable low input and high through epigenomics profiling. First, I examined RNA polymerase II transcriptional regulation with microfluidic chromatin immunoprecipitation followed by next generation sequencing (ChIP-seq) assays. Second, I probed the temporal dynamics in the DNA methylome during cancer development using a transgenic mouse model with microfluidic methylated DNA immunoprecipitation followed by next generation sequencing (MeDIP-seq) assays. Third, I explored negative enrichment of circulating tumor cells (CTCs) followed by microfluidic ChIP-seq technology for studying temporal dynamic histone modification (H3K4me3) of patient-derived tumor xenograft on an immunodeficient mouse model during the course of cancer metastasis. In the first study, I adapted microfluidic ChIP-seq devices to achieve ultrahigh sensitivity to study Pol2 transcriptional regulation from scarce cell samples. I dramatically increased the assay sensitivity to an unprecedented level (~50 K cells for pol2 ChIP-seq). Importantly, this is three orders of magnitude more sensitive than the prevailing pol2 ChIP-seq assays. I showed that MNase digestion provided better ChIP-seq signal than sonication, and two-steps fixation with MNase digestion provided the best ChIP-seq quality followed by one-step fixation with MNase digestion, and lastly, no fixation with MNase digestion. In the second study, I probed dynamic epigenomic changes during tumorigenesis using mice often require profiling epigenomes using a tiny quantity of tissue samples. Conventional epigenomic tests do not support such analysis due to the large amount of materials required by these assays. In this study, I developed an ultrasensitive microfluidics-based methylated DNA immunoprecipitation followed by next-generation sequencing (MeDIP-seq) technology for profiling methylomes using as little as 0.5 ng DNA (or ~100 cells) with 1.5 h on-chip process for immunoprecipitation. This technology enabled me to examine genome-wide DNA methylation in a C3(1)/SV40 T-antigen transgenic mouse model during different stages of mammary cancer development. Using this data, I identified differentially methylated regions and their associated genes in different periods of cancer development. Interestingly, the results showed that methylomic features are dynamic and change with tumor developmental stage. In the last study, I developed a negative enrichment of CTCs followed by ultrasensitive microfluidic ChIP-seq technology for profiling histone modification (H3K4Me3) of CTCs to resolve the technical challenges associated with CTC isolation and difficulties related with tools for profiling whole genome histone modification on tiny cell samples. / Ph. D. / The human genome has been sequenced and completed over a decade ago. The information provided by the genomic map inspired numerous studies on genetic variations and their roles in diseases. However, genomic information alone is not always sufficient to explain important biological processes. Gene activation and expression are not only associated with alteration in the DNA sequence, but also affected by other changes to DNA and histones. Epigenetics refers to the molecular mechanisms that affect gene expression and phenotypes without involving changes in the DNA sequence. For example, the DNA can get methylated, the histone protein that is wrapped around by DNA can also get methylated or acetylatied, and transcription factors can bind to different part of DNA. All of these can affect gene expression without alter the DNA sequences. Epigenetic changes occur throughout all stages of cell development or in response to environmental cues. They change transcription patterns in a tissue/cell-specific fashion. For example, transcriptional silencing of tumor-suppressor genes by DNA methylation plays an important role in cancer development. Therefore, understanding of epigenetic regulations will help to improve various aspects of biomedicine. For instance, personalized medicine can be vi tailored based on epigenetic profile of certain patient to specifically control gene expression in the disease treatment. However, the technology for profiling epigenetic modifications, i.e. Chromatin Immunoprecipitation (ChIP), suffers from serious limitations. The key limitation is the sensitivity of the assay. Conventional assay requires a large number of cells (>10⁶ cells per ChIP). This is feasible when using cell lines. However, such requirement has become a major challenge when primary cells are used because very limited amounts of samples can be generated from lab animals or patients. Population heterogeneity information may also be lost when a large cell number is used. In this project, we developed an automated ultrasensitive microfluidic chromatin/DNA immunoprecipitation followed by next-generation sequencing (ChIP/MeDIP-Seq) technology for profiling epigenetic modifications (e.g., histone modifications, transcriptional regulations, and DNA methylation). We extensively optimized design parameters for each and every step of ChIP/MeDIP (e.g. sonication/crosslinking time, antibody concentration, washing conditions) in order to reach highest sensitivity of 0.1 ng DNA (or ~50-100 cells) as starting material for IP, which is roughly 4-5 orders of magnitude higher than the prevailing protocol and 2-3 orders of magnitude higher than the-state-of-the-art(~50 ng). With such sensitivity, we were able to study temporal dynamics in the DNA methylomes during the various stages of mammary cancer development from a transgenic mouse mode. We were able to investigate transcriptional regulation of RNA polymerase II from scarce cell samples. We were also able to study histone modification (H3K4Me3) of circulating tumor cells during cancer metastasis.

Hormone-dependent cancers (HDCs), such as those of the prostate, ovary, breast and endometrium, share characteristics that indicate similar underlying mechanisms of carcinogenesis. Through steroid hormone signalling on "down-stream" target genes, the growth, development and progression of HDCs are regulated. One such family of target genes, highly expressed in HDCs and regulated by steroid hormones, are the tissue kallikreins (KLKs). The KLKs are a multigene family of serine proteases involved in physiological processes such as blood pressure regulation, inflammation, and tumour development and progression via the hydrolysis of specific substrates. Although the KLK gene family is clearly implicated in tumourigenesis, the precise roles played by these genes are largely unknown. Additionally, except for the androgen-responsive genes, KLK2 and KLK3, the mechanisms underlying their hormonal regulation in HDCs are yet to be identified. The initial focus of this thesis was to examine the regulation of the kallikreins, KLK1 and KLK4, by estradiol and progesterone in endometrial and breast cancer cell lines. From these studies, progesterone clearly regulated KLK4 expression in T47D cells and therefore, the focus of the remaining studies was to further examine this regulation at the transcriptional level. An overview of the results obtained is detailed below. Human K1 and hK4 protein levels were increased by 10 nmol/L estradiol benzoate, progesterone, or a combination of the two, over 48 hours in the endometrial cancer cell line, KLE. However, these same treatments resulted in no change in KLK1 gene or hK1 protein levels in the endometrial cancer cell lines, HEC1A or HEC1B (only hK1 analysed). Progesterone treatment (0-100 nmol/L) over 24 hours resulted in a clear increase in KLK4 mRNA at the 10 nmol/L dose in the breast cancer cell line, T47D. Additionally, treatment of T47D cells with 10 nmol/L progesterone over 0-48 hr, resulted in the rapid expression of the hK4 protein at 2 hr which was sustained for 24 hr. Further analysis of this latter progesterone regulation with the antiprogesterone, RU486, over 24 hours, resulted in an observable decrease in hK4 levels at 1 µmol/L RU486. Although the estrogen and progesterone regulation of the hK1 protein was not further analysed, the data obtained for hK4 regulation in T47D cell lines, supported the premise that this gene was progesterone-responsive. The rapid expression of hK4 protein by progesterone at two hours suggests that KLK4 transcription is directly coupled to progesterone regulation, perhaps through progesterone receptor (PR) binding to progesterone-responsive regions within the KLK4 promoter or far "up-stream" regions. Thus, the following further studies were performed. To test this hypothesis, the transcription initiation site (TIS) and 5' flanking regions of the KLK4 gene in T47D cells were interrogated. Primer extension and 5' RACE identified the TIS 78 bp 5' of the putative ATG site for translation as identified by Korkmaz et al. (2001). This KLK4 gene transcript consists of only four exons, and thus excludes the pre/pro signal peptide. Although a TATA-box is not present within -25 to -30 bp 5' of the identified TIS, a number of consensus binding motifs for Sp1 and estrogen receptor half-sites were identified. It is possible that the Sp1 sites are involved in the basal levels of transcription for this gene. Additionally, a putative progesterone response element (PRE) was identified in the far "up-stream" regions of the KLK4 gene. Basal levels of transcription were observed within the KLK4 proximal promoter region when coupled to a luciferase reporter gene and transfected into T47D cell lines. Additionally, the KLK4 proximal promoter region did not induce the luciferase reporter gene expression when progesterone was added to the system, however, estradiol was inhibitory for luciferase gene expression. This suggests that the proximal promoter region of the KLK4 gene could contain functional EREs but not PREs. In keeping with this hypothesis, some ER half-sites were identified, but PR sites were not obvious within this region. The identified PRE in the far "up-stream" region of the KLK4 gene assembled the progesterone receptor in vitro, and in vivo, as assessed by electromobility shift assays and chromatin immunoprecipitation assays (EMSAs and ChIPs), respectively. The binding of the PR to the KLK4 PRE was successfully competed out by a PR antibody and not by an androgen receptor antibody, and thus confirms the specificity of the KLK4 PRE-PR complex. Additionally, the PR was recruited and assembled onto and off the progesterone-responsive KLK4 region in a cyclic fashion. Thus, these data strongly suggest that the PR represents one of the core components of a transcription complex for the KLK4 gene, and presumably also contributes to the expression of this gene. Moreover, these data suggest a functional coordination between the PR and the KLK4 progesterone-responsive region in T47D cells, and thus, provide a model system to further study these events in vivo.