Thesis: S.M., Massachusetts Institute of Technology, Computational and Systems Biology Program, 2014. / Cataloged from PDF version of thesis. / Includes bibliographical references (page 29). / Caspases are a 12-member family of human proteases that regulate apoptosis and inflammation. They serve as key effectors downstream of diverse signaling receptors and shape cell fate. Inflammatory caspases mediate the proteolytic processing of inflammatory cytokines and are essential in maintaining immune function, but also lead to disease when deregulated. In order to examine the activity of inflammatory caspases, we generated 2 inflammatory caspase reporters: a fluorescence resonance energy transfer (FRET) inflammatory caspase activity reporter as well as a fluorescent translocation inflammatory caspase reporter. These reporters were then used to study inflammatory caspase activity in vitro using recombinant caspases and in vivo using a simplified cell culture model. The inflammatory caspase activity reporters have the potential to capture inflammatory caspase activation under a variety of stimuli. They also have several advantages compared to existing methods: they are non-destructive and can be used for live single cell measurements; they do not require the addition of exogenous chemicals or cofactors; and they do not covalently modify the inflammatory caspases. Inflammatory caspase activation is a rapid, asynchronous process, and detecting the activity of the mature inflammatory caspase molecules is made difficult due to the short half-life of the enzyme. The reporters we have developed can fill this need. / by Arshed Al-Obeidi. / S.M.

Thesis (Ph. D.)--Massachusetts Institute of Technology, Computational and Systems Biology Program, 2010. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Cataloged from student-submitted PDF version of thesis. / Includes bibliographical references (p. 159-176). / microRNAs (miRNAs) are ~22-nt long short RNAs that regulate gene expression in organisms ranging from plants to animals. In mammals, miRNAs post-transcriptionally repress gene expression by primarily binding to the 3' untranslated region (3' UTR) of target mRNAs. Although hundreds of miRNAs have been discovered, targets of most miRNAs and the method by which they affect their biological function remain elusive. To better understand the role of miRNAs in fundamental cellular processes, we characterized enriched miRNA populations in three distinct murine developmental programs, T lymphocytes, embryonic stem cells, and the placenta. We started exploring the role of miRNAs in T lymphocytes by globally characterizing short RNA expression during key developmental stages of T lymphocytes. Our results showed that a distinct set of miRNAs is enriched in each stage. In particular, miR-181 is elevated at the double positive (DP) stage, when thymocytes expressing both CD4 and CD8 undergo positive and negative selection. We found that miR-181 can repress the expression of Bcl-2, CD69, and the T cell receptor, all of which are involved in positive selection. Analysis of short RNAs in T lymphocytes also revealed a novel miRNA cluster, the Sfmbt2 miRNA cluster, named as such since it maps to an intron of the Sfmbt2 gene, a Polycomb Group gene. Instead of studying this cluster in T lymphocytes, we decided to use embryonic stem (ES) cells as this cluster is also expressed in ES cells and the cells are more conducive to lab experimentation. This cluster contains several miRNA families, and we addressed the function of one miRNA family, miR-467a, as it shares target specificity with other highly abundant miRNAs in ES cells. Gain and loss of function assays showed that this family of miRNAs can promote cell survival by advancing the G1 to S phase transition. In addition, they target certain proapoptotic factors to buffer ES cells from apoptosis, especially in the context of genotoxic stress. The Sfmbt2 cluster is a mouse-specific miRNA cluster, and individual members have been uniquely amplified in the Sfmbt2 locus. We developed a method to explore the impact of species-specific miRNAs on the evolution of 3' UTRs, and found that target sites of many miRNAs show positive selection. In particular, mouse target sites have evolved to specifically gain binding sites (mouse-specific targets) for some Sfmbt2 miRNAs, several of which are enriched in the placenta. These mouse-specific targets are enriched in pathways regulating cell survival, implicating the Sfmbt2 miRNA cluster as a possible promoter to placental growth. Our studies in T lymphocytes, ES cells and the placenta have revealed important roles of miRNAs in shaping 3' UTR evolution, and mammalian development. Several novel miRNA targets we uncovered are important regulators of differentiation, cell cycle, and apoptosis. Understanding their functions will not only shed light on their roles in normal physiology, but also generate useful insights that can be applied to cancer and reprogramming. / by Grace Xinying Zheng. / Ph.D.

Thesis: Ph. D., Massachusetts Institute of Technology, Computational and Systems Biology Program, 2015. / Cataloged from PDF version of thesis. / Includes bibliographical references. / Mammals contain a wide array of cell types with distinct functions, yet nearly all cell types have the same genomic DNA. How the genetic instructions in DNA are selectively interpreted by cells to specify various cellular functions is a fundamental question in biology. This thesis work describes two genome-wide studies designed to study how transcriptional control of gene expression programs defines cell identity. Recent studies suggest that a small number of transcription factors, called "master" transcription factors, dominate the control of gene expression programs. These master transcription factors and the transcriptional regulatory circuitry they produce, however, are not known for all cell types. Ectopic expression of these factors can, in principle, direct transdifferentiation of readily available cells into medically relevant cell types for applications in regenerative medicine. Limited knowledge of these factors is a roadblock to generation of many medically relevant cell types. Chapter 2 presents a study in which a novel computational approach was undertaken to generate an atlas of candidate master transcriptional factors for 100+ human tissue/cell types. The candidate master transcription factors in retinal pigment epithelial (RPE) cells were then used to guide the investigation of the regulatory circuitry of RPE cells and to reprogram human fibroblasts into functional RPE-like cells. Master transcription factors define cell-type-specific gene expression through binding to enhancer elements in the genome. These enhancer-bound transcription factors regulate genes by contacting target gene promoters via the formation of DNA loops. It is becoming increasingly clear that transcription factors operate and regulate gene expression within a larger three-dimensional (3D) chromatin architecture, but these structures and their functions are poorly understood. Chapter 3 presents a study in which Cohesin ChIA-PET data was generated to identify the local chromosomal structures at both active and repressed genes across the genome in embryonic stem cells. The results led to the discovery of functional insulated neighborhood structures that are formed by two CTCF interaction sites occupied by Cohesin. The integrity of these looped structures contributes to the transcriptional control of super-enhancer-driven active genes and repressed genes encoding lineage-specifying developmental regulators. / by Zi Peng Fan. / Ph. D.

Thesis: Ph. D., Massachusetts Institute of Technology, Computational and Systems Biology Program, 2017. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Cataloged from student-submitted PDF version of thesis. / Includes bibliographical references (pages 129-140). / Microbial genomes show high plasticity due to horizontal transfer, large community sizes, and rapid growth paired with adaptive mutations. Despite this mutability of gene content, most studies of microbial communities still rely on bulk, single-gene amplicon sequencing. In this thesis, I present methods that interrogate the gene content of single cells derived from complex natural communities. In the first project, I present a novel molecular biology method to link a bacterial functional gene to its host species with single-cell resolution. This high-throughput protocol is applied to assess the distribution of anaerobic respiration genes in a lake ecosystem. In the second project, I demonstrate extensions of this methodology to link genes between spatially proximal microbial cells, and apply this approach to probe the spatial organization of human dental plaque using DNA sequencing. In the final project, I completed whole-genome sequencing of environmental isolates derived from single, cultivable cells and employ mutational and horizontal transfer analysis to demonstrate adaptation to harsh environmental conditions in contaminated groundwater. These projects demonstrate the rich information stored within each microbial genome and the impact of spatial distribution in the environment. Each effort also contributes or highlights new molecular biology techniques to generate genomic data from individual microbial cells. / by Sarah Jean Spencer. / Ph. D.

Thesis (S.M.)--Massachusetts Institute of Technology, Computational and Systems Biology Program, 2012. / Cataloged from PDF version of thesis. / Includes bibliographical references (p. 47-53). / Tyrosine kinase inhibitors have significant promise in the fight to develop agents that can target cancer in a tumor-specific manner. A number of drugs have been and are currently in development to inhibit specific kinases that can mediate uncontrolled proliferation; however, an unfortunate eventuality for most patients receiving these treatments is the development of resistance that renders these drugs almost completely ineffective. While a number of mechanisms can evolve within a tumor to mitigate effects of kinase inhibitors, we sought to uncover what changes are occurring in the tyrosine phosphorylation network at both short timescales (minutes to 72 hours) and long timescales (120 hours+) that can be playing a role in helping a tumor become resistant to driver-kinase inhibition. It is our hypothesis that specific feedback networks are able to detect and overcome driver kinase inhibition through activation of potential other pathways, which can go on to mediate a longer term resistance phenotype. In order to probe dynamics in the tyrosine phosphorylation network, we employed mass spectrometry to analyze peptides derived from six non-small cell lung cancer cell lines that we classify as either EGFR+ or EML4-ALK+. From both mass spectrometry data and growth assays, we identified an unintuitive boost in signaling and growth in response to low inhibitor concentrations, suggestive of a cellular mechanism that is adaptive to driver kinase inhibition. Studies of EML4-ALK driven H3122 cells showed that this short-term response is not the same as the known long-term resistance mechanism to ALK inhibition, leading support to the notion that the short-term "adaptive response" may be a novel type of mechanism to aid tumor adaptation to targeted therapies. In an effort to better probe signaling events occurring downstream of the phosphotyrosine network, a new pull down technique for mass spectrometry using 14-3-3 protein against phosphoserine and phosphothreonine peptides is described. The results of these studies open up many potential avenues for further exploration into the immediate and long-term signaling response of cancer to targeted therapies. / by Adrian Joseph Randall. / S.M.

Thesis: Ph. D., Massachusetts Institute of Technology, Computational and Systems Biology Program, 2014. / This electronic version was submitted by the student author. The certified thesis is available in the Institute Archives and Special Collections. / Cataloged from student-submitted PDF version of thesis. / Includes bibliographical references. / In the post-genomic era, greater emphasis has been placed on understanding the function of genes at the systems level. To meet these needs, biologists are creating larger, and increasingly complex datasets. In recent years, high-content screening (HCS) using RNA interference (RNAi) or other perturbation techniques in combination with automated microscopy has emerged as a promising investigative tool to explore intricate biological processes. Image-based HC screens produce massive hyperdimensional data sets. To identify novel components of the DNA damage response (DDR) after ionizing radiation, we recently performed an image-based HC RNAi screen in an osteosarcoma cell line. Robust univariate hit identication methods and manual network analysis identied an isoform of BRD4, a bromodomain and extra-terminal domain family member, as an endogenous inhibitor of DDR signaling. However, despite the plethora of data generated from our and other HC screens, little progress has been made in analyzing HC data using multivariate computational methods that exploit the full richness of hyperdimensional data and identify more than just the most salient knockdown phenotypes to gain a detailed understanding of how gene products cooperate to regulate complex cellular processes. We developed a novel multivariate method using logistic regression models and least absolute shrinkage and selection operator regularization for analyzing hyperdimensional HC data. We applied this method to our HC screen to identify genes that exhibit subtle but consistent phenotypic changes upon knockdown that would have been missed by conventional univariate hit identication approaches. Our method automatically selects the most predictive features at the most predictive time points to facilitate the more ecient design of follow-up experiments and puts the identied hits in a network context using the Prize-Collecting Steiner Tree algorithm. This method offers superior performance over the current gold standard for the analysis of HC RNAi screens. A surprising finding from our analysis is that training sets of genes involved in complex biological phenomena used to train predictive models must be broken down into functionally coherent subsets in order to enhance new gene discovery. Additionally, we found that in the case of RNAi screening, statistical cell-to-cell variation in phenotypic responses in a well of cells targeted by a single shRNA is an important predictor of gene dependent events. / by Jonathan Rameseder. / Ph. D.

Thesis: Ph. D., Massachusetts Institute of Technology, Computational and Systems Biology Program, 2014. / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 167-187). / Posttranscriptional regulation of mRNA by RNA-binding proteins plays key roles in regulating the transcriptome over the course of development, between tissues and in disease states. The specific interactions between mRNA and protein are controlled by the proteins' inherent affinities for different RNA sequences as well as other features such as translation and RNA structure which affect the accessibility of mRNA. The stabilities of mRNA transcripts are regulated by nonsense-mediated mRNA decay (NMD), a quality control degradation pathway. In this thesis, I present a novel method for high throughput characterization of the binding affinities of proteins for mRNA sequences and an integrative analysis of NMD using deep sequencing data. This thesis describes RNA Bind-n-Seq (RBNS), which comprehensively characterizes the sequence and structural specificity of RNA binding proteins (RBPs), and application to the developmentally-regulated splicing factors RBFOX2, MBNL1 and CELF1/CUGBP1. For each factor, the canonical motifs are recovered as well as additional near-optimal binding motifs. RNA secondary structure inhibits binding of RBFOX2 and CELF1, while MBNL1 favors unpaired Us but tolerates C/G pairing in UGC-containing motifs. In a project investigating how NMD shapes the embryonic transcriptome, this thesis presents integrated genome-wide analyses of UPF1 binding locations, NMD-regulated gene expression, and translation in murine embryonic stem cells (mESCs). Over 200 direct UPF1 binding targets are identified using crosslinking/immunoprecipitation-sequencing (CLIP-seq). Results from ribosome foot printing show that actively translated upstream open reading frames (uORFs) are enriched in transcription factor mRNAs and predict mRNA repression by NMD, while poorly translated mRNAs escape repression. / by Alexander De Jong Robertson. / Ph. D.

Thesis: Ph. D., Massachusetts Institute of Technology, Computational and Systems Biology Program, 2014. / Cataloged from PDF version of thesis. / Includes bibliographical references. / Transmission of information from DNA to RNA to protein underlies the core of modem life forms. The advance in sequencing and genetic technologies has revolutionized the study of molecular biology, genetics and developmental biology enabling delineation of biological processes in unprecedented details. Through the study of epigenetics and posttranscriptional regulation of gene expression by high-throughput sequencing technologies in several biological processes, namely embryonic stem cells, somatic reprogramming, erythroid differentiation, epithelial-mesenchymal transition and cancer metastasis, this thesis work has identified novel players and regulatory mechanisms underlying these developmental processes and diseases. Furthermore, an attempt to engineer CRISPRzymes - protein fusions of RNA-guided DNA binding dCas9 - will enable experiments to directly test biological processes at defined genomic loci and expands the toolbox for synthetic biology and potentially opens up opportunities for novel therapeutics. / by Wu Albert Cheng. / Ph. D.

Thesis: Ph. D., Massachusetts Institute of Technology, Computational and Systems Biology Program, 2016. / Cataloged from PDF version of thesis. / Includes bibliographical references. / Mammalian spermatogenesis includes two types of cell divisions. First, germ cells undergo transit-amplifying mitotic divisions, which enable prodigious output of mature spermatozoa. Second, they undergo reductive meiotic divisions to produce haploid gametes. In this thesis, I examine gene expression and regulation during the mitotic and meiotic phases of spermatogenesis. Chapter 2 describes how RA-STRA8 signaling regulates two key transitions: spermatogonial differentiation, which begins the transit-amplifying mitotic divisions, and meiotic initiation, which ends them. First, in mice lacking the RA (retinoic acid) target gene Stra8, undifferentiated spermatogonia accumulated; thus, Stra8 promotes spermatogonial differentiation as well as meiotic initiation. Second, injection of RA into wild-type males induced precocious spermatogonial differentiation and meiotic initiation; thus, RA acts instructively on germ cells at both transitions. Finally, competencies of germ cells to undergo spermatogonial differentiation or meiotic initiation in response to RA were found to be distinct and periodic. Chapter 3 describes a novel method for isolating precise populations of mitotic and meiotic germ cells from the mouse testis. We first synchronize germ cell development in vivo, and perform histological staging to verify synchronization. We then separate these germ cells from contaminating somatic and stem cells by FACS, to achieve ~90% purity of each distinct germ cell type, from the stem cell pool through mid/late meiotic prophase. Utilizing this "3S" method (synchronize, stage, and sort), we can robustly and efficiently separate germ cell types that were previously challenging or impossible to distinguish, with sufficient yield for transcriptomic and epigenetic studies. Chapter 4 presents a systematic comparison of the male and female gene expression programs of meiotic prophase. We performed transcriptional profiling of postnatal testes synchronized in precise stages of meiotic prophase, and compared to the same stages in the fetal ovary. We identified 260 genes up-regulated during both male and female prophase; this shared gene set represents a core meiotic program, composed of known and potential novel meiotic players. We also identified over two thousand genes that are up-regulated during meiotic prophase specifically in the male. These comprise both a male-specific meiotic program, and a preparatory program for cellular differentiation of spermatozoa. / by Katherine A. Romer. / Ph. D.

Thesis: Ph. D., Massachusetts Institute of Technology, Computational and Systems Biology Program, 2014. / Cataloged from PDF version of thesis. / Includes bibliographical references / The vast majority of the mammalian genome does not encode proteins. Only 2% of the genome is exonic, yet recent deep survey of human transcripitome suggested that 75% of the genome is transcribed, including half of the intergenic regions. Such pervasive transcription typically leads to short-lived, low-copy number noncoding RNAs (ncRNAs). We are starting to understand the biogenesis and mechanisms regulating the noncoding transcription. However, it is still unclear what's the functional impact of pervasive transcription and the ncRNAs at the level of the'genome, the cell, and the organism. A large fraction of ncRNAs in cells is generated by divergent transcription that occurs at the majority of mammalian gene promoters. RNA polymerases transcribe divergently on opposite strands, producing precursor mRNAs (pre-mRNAs) on one side and promoter upstream antisense RNAs (uaRNAs) on the other side. Like typical products of pervasive transcription, uaRNAs are relatively short and unstable as compared to pre-mRNAs, suggesting there are mechanisms suppressing uaRNA transcription and enforcing promoter directionality. We describe the Ul-PAS axis, a mechanism that enhances gene transcription but suppresses noncoding transcription. Two RNA processing signals, the Ul signal, or 5' splice site sequences recognized by Ul snRNP during splicing, and polyadenylation signal (PAS), differentially mark the two sides of gene transcription start site (TSS), ensuring the generation of full-length mRNA but inducing early termination of uaRNAs. The Ul-PAS axis also suppresses pervasive transcription on the antisense strand of genes, as well as intergenic transcription. Transcription is a mutagenic process that could accelerate evolution. We uncover a link between pervasive transcription and genome evolution. Specifically, transcription-induced mutational bias in germ cells could strengthen the Ul-PAS axis, which in turn enhances transcription, thus forming a positive feedback loop, which eventually drives new gene origination, and facilitates genome rearrangements. Tools to directly interfere with transcription with specificity are necessary to understand the function of noncoding transcription, especially when the RNA product is rapidly degraded or nonfunctional. The newly emerged CRISPR-Cas9 system provides the opportunity to target any desired locus. We comprehensively characterize the binding specificity of Cas9 in the mouse genome. We find that Cas9 specificity varies dramatically but in a predictable manner, depending on the seed sequence and chromatin accessibility. Our results will facilitate Cas9 target design and enable genome manipulation with high precision. / by Xuebing Wu. / Ph. D.