published_or_final_version / abstract / Microbiology / Doctoral / Doctor of Philosophy

Cellular proliferation is a key characteristic of hematopoietic stem and progenitor cells (HSC/HPCs) that allows for the production of all blood cell lineages during an individuals lifetime. While this feature of stem cells is strictly regulated during steadystate and stress hematopoiesis, it also contributes to the development of myeloproliferative disorders, such as chronic myelogenous leukemia, essential thrombocythemia, and polycythemia vera. It should come as no surprise then, that common treatments for these diseases often target the proliferative nature of the dysfunctional HSC/HPCs. Thus, the identification of molecular determinants of cell cycle regulation associated with these disorders could serve as targets for novel therapies. Using the hematopoietic system of the inbred mouse strains, C57BL/6J (B6) and DBA/2J (D2), it was found that the HSC/HPCs of the long-lived B6 mouse strain were less susceptible to the cytostatic agent hydroxyurea (HU) than the short-lived D2 mouse strain. A quantitative trait locus (QTL) analysis revealed a region of proximal chromosome 7 that regulates this response to HU. Congenic mouse strains were generated and phenotypic analysis confirmed that the B6 and D2 loci confer a low and high sensitivity of the HSC/HPCs to HU, respectively. We then showed that while this response of the HSC/HPCs to HU is independent of their cell cycle status, the B6 allele of this QTL confers a proliferative advantage to bone marrow cells after bone marrow transplantation. Having shown that proximal chromosome 7 regulates the response of HSC/HPCs to HU, we found it necessary to characterize the gene and protein expression profiles in order to identify the responsible candidate genes. We first analyzed mRNA expression profiles of HPCs from the parental and congenic mouse strains using gene microarrays and found that four genes within the congenic interval were differentially expressed. Real-time PCR confirmed that the expression profile of only one gene, Ndufa3, is significantly different in HPCs of B6 and D2 mice. Concurrently, we assessed the protein expression profiles of HPC-enriched mononuclear cells. Significant differences were found between the cytoplasmic and nuclear fractions of both strains, with a skewing of protein expression towards the D2 congenic strain.

Signal Track Analytical Research Tool (START), is a parallel system for analyzing large-scale genomic data. Currently, genomic data analyses are usually performed by using custom scripts developed by individual research groups, and/or by the integrated use of multiple existing tools (such as BEDTools and Galaxy). The goals of START are 1) to provide a single tool that supports a wide spectrum of genomic data analyses that are commonly done by analysts; and 2) to greatly simplify these analysis tasks by means of a simple declarative language (STQL) with which users only need to specify what they want to do, rather than the detailed computational steps as to how the analysis task should be performed. START consists of four major components: 1) A declarative language called Signal Track Query Language (STQL), which is a SQL-like language we specifically designed to suit the needs for analyzing genomic signal tracks. 2) A STQL processing system built on top of a large-scale distributed architecture. The system is based on the Hadoop distributed storage and the MapReduce Big Data processing framework. It processes each user query using multiple machines in parallel. 3) A simple and user-friendly web site that helps users construct and execute queries, upload/download compressed data files in various formats, man-age stored data, queries and analysis results, and share queries with other users. It also provides a complete help system, detailed specification of STQL, and a large number of sample queries for users to learn STQL and try START easily. Private files and queries are not accessible by other users. 4) A repository of public data popularly used for large-scale genomic data analysis, including data from ENCODE and Roadmap Epigenomics, that users can use in their analyses. / published_or_final_version / Computer Science / Doctoral / Doctor of Philosophy

Hereditary bacterial symbiosis is a common mechanism by which eukaryotic hosts can acquire traits beneficial for their fitness. Many insects have symbiotic associations with bacteria that trace back millions of years, whose function and evolution are well characterized. Insects can also possess more recently derived symbionts that are closely related to free-living bacteria, and often play a role in host defense. Serratia symbiotica is a recently derived symbiont that infects aphids and provides protection against heat stress, and possibly also plays a nutritional role. Many aspects of the biology of recent symbionts are less well studied, including the diversity of functional roles and evolution among hosts for single lineages of symbionts, the molecular mechanisms that contribute to defense, the early stages of symbiont genome evolution, and interactions with hosts. This dissertation focuses upon S. symbiotica to contribute research addressing each of these themes. Functional studies revealed that S. symbiotica lysis during heat-shock is correlated with protection of the nutritional symbiont Buchnera , and that S. symbiotica has a large effect upon aphid metabolite pools. Despite this large metabolic effect, S. symbiotica does not seem to dramatically influence expression of aphid genes, including those involved in immunity. Analysis of the evolution of S. symbiotica lineages in different aphid hosts revealed this symbiont is common in the aphid subfamily Lachninae, but did not support the obligate nutritional role hypothesized in the literature for this group. Finally, comparison of the S. symbiotica genome to close free-living relatives revealed a genome undergoing massive decay, and provided a rare opportunity to examine the evolution of a recently acquired symbiont.

<p>Chlamydia trachomatis, a pathogen responsible for major diseases of significant clinical and public health importance, remains poorly characterized because of its intractability to molecular genetic manipulation. The development of a system(s) for genetic analysis would significantly accelerate our ability to identify genes that enable Chlamydia to establish infection, survive within its host, and cause disease. This thesis describes two methods used to assess gene function in Chlamydia and to provide insights into its biology and pathogenesis. The first method described is based on specific inhibitors and is used to probe the role of lipooligosaccharide (LOS), a main lipid components of bacterial outer membranes. Using this approach, we show that small molecule inhibitors of LpxC [UDP-3-O-(R-3-hydroxymyristoyl)-GlcNAc deacetylase], the enzyme that catalyzes the first committed step in the biosynthesis of lipid A, blocks the synthesis of LOS in C. trachomatis. In the absence of LOS, Chlamydia remains viable and establishes a pathogenic vacuole ("inclusion") that supports robust bacterial replication. However, bacteria grown under these conditions were no longer infectious. In the presence of LpxC inhibitors, replicative reticulate bodies accumulated in enlarged inclusions but failed to express selected late-stage proteins and transition to elementary bodies, a Chlamydia developmental form that is required for invasion of mammalian cells. These findings suggest the presence of an outer membrane quality control system that regulates Chlamydia developmental transition to infectious elementary bodies and highlights the potential application of LpxC inhibitors as unique class of anti-chlamydial agents.</p><p> The second part of this thesis describes the development of a system with which to perform forward genetics in C. trachomatis. Forward genetics approaches set out to identify the gene or set of genes that contributes to a specific biological process and usually entails generating random mutations in a large number of organisms, isolating mutants with an aberrant phenotype, and identifying the alleles associated with the mutant phenotype. In this approach, chemical mutagenesis is coupled with whole genome sequencing (WGS) and a system for DNA exchange within infected cells to generate Chlamydia mutants with distinct phenotypes, map the underlying genetic lesions, and generate isogenic strains. We identified mutants with altered glycogen metabolism, including an attenuated strain defective for Type II secretion. The coupling of chemically induced gene variations and WGS to establish genotype-phenotype associations should be broadly applicable to the growing list of microorganisms intractable to traditional genetic mutational analysis.</p> / Dissertation

Transcriptomic analyses from across eukaryotes indicate that most of the genome is transcribed at some point in the developmental trajectory of an organism. One class of these transcripts is termed long intergenic noncoding RNAs (lincRNAs). Recently, attention has focused on understanding the evolutionary dynamics of lincRNAs, particularly their conservation within genomes. Here, we take a comparative genomic and phylogenetic approach to uncover factors influencing lincRNA emergence and persistence in the plant family Brassicaceae, to which Arabidopsis thaliana belongs. We searched 10 genomes across the family for evidence of >5000 lincRNA loci from A. thaliana. From loci conserved in the genomes of multiple species, we built alignments and inferred phylogeny. We then used gene tree/species tree reconciliation to examine the duplication history and timing of emergence of these loci. Emergence of lincRNA loci appears to be linked to local duplication events, but, surprisingly, not whole genome duplication events (WGD), or transposable elements. Interestingly, WGD events are associated with the loss of loci for species having undergone relatively recent polyploidy. Lastly, we identify 1180 loci of the 6480 previously annotated A. thaliana lincRNAs (18%) with elevated levels of conservation. These conserved lincRNAs show higher expression, and are enriched for stress-responsiveness and cis-regulatory motifs known as conserved noncoding sequences (CNSs). These data highlight potential functional pathways and suggest that CNSs may regulate neighboring genes at both the genomic and transcriptomic level. In sum, we provide insight into processes that may influence lincRNA diversification by providing an evolutionary context for previously annotated lincRNAs.

Functional genomics uses high throughput genome-wide technologies to investigate the functional consequences of genetic variants on gene expression and protein products. In the context of disease, using integrative functional genomic approaches to understand the genetic variation which underlies disease susceptibility and aetiology may contribute to better diagnosis, treatment and prevention. This thesis investigated genetic determinants of variation in response to infection by applying a functional genomics approach to investigate three clinical cohorts: patients with severe sepsis, an influenza challenge study and patients with common variable immune deficiency disorders. The transcriptomic response to severe sepsis is reported here for the largest known adult severe sepsis community acquired pneumonia cohort. Two clusters within the cohort, based on gene expression signatures, which have different survival rates and identify individuals with distinct immune responses to sepsis, highlight the value of functional genomics for identifying heterogeneity within patient cohorts. This was further complemented by the resolution of gene expression signatures in healthy individuals following influenza challenge which identified individuals with moderate to severe disease. Shared gene expression signatures between the cohorts highlighted components of the immune response to viral infection important across multiple clinical settings and may assist with the identification of viral infections in the sepsis patients. Gene expression was mapped as a quantitative trait (eQTL). Comparison to data sets for healthy individuals and from innate immune stimulated cells identified regulatory variants specific to the context of sepsis. Whole genome sequencing for a cohort of patients with common variable immune deficiency disorders was performed. This identified novel variants and pathways which may play a role in the underlying immunopathogenesis of disease. Integration with RNA-seq for a small number of patients allowed prioritisation of non-coding variants based on evidence of altered gene expression. Comparison to the sepsis cohort analysis identified key immune related genes involved in rare and common responses to bacterial infection. This thesis has demonstrated the value of integrating multiple functional genomic techniques to further our understanding of the mechanisms underlying variation in the response to infection.

Multigenic diagnostic and prognostic tools can be derived for ordinal clinical outcomes using data from high-throughput genomic experiments. A challenge in this setting is that the number of predictors is much greater than the sample size, so traditional ordinal response modeling techniques must be exchanged for more specialized approaches. Existing methods perform well on some datasets, but there is room for improvement in terms of variable selection and predictive accuracy. Therefore, we extended an impressive binary response modeling technique, Feature Augmentation via Nonparametrics and Selection, to the ordinal response setting. Through simulation studies and analyses of high-throughput genomic datasets, we showed that our Ordinal FANS method is sensitive and specific when discriminating between important and unimportant features from the high-dimensional feature space and is highly competitive in terms of predictive accuracy. Discrete survival time is another example of an ordinal response. For many illnesses and chronic conditions, it is impossible to record the precise date and time of disease onset or relapse. Further, the HIPPA Privacy Rule prevents recording of protected health information which includes all elements of dates (except year), so in the absence of a “limited dataset,” date of diagnosis or date of death are not available for calculating overall survival. Thus, we developed a method that is suitable for modeling high-dimensional discrete survival time data and assessed its performance by conducting a simulation study and by predicting the discrete survival times of acute myeloid leukemia patients using a high-dimensional dataset.

Human clinical genome-wide association studies (GWAS) have helped identify disease trait and pharmacogenomic loci without the need for biological under- standing. Molecular GWAS - associating genetic variation with traits such as gene expression - have been slow to fill the mechanistic gaps. While tissue specificity, lack of DNA resolution, and the need for better data integration are no dou bt important bottlenecks in molecular GWAS, there is also a very poor general understanding of which molecular phenotypes are important and how best to model them. Added to this is the clear need for a greater understanding of the strengths and weaknesses facing in vitro (and ex vivo) models as hypoth- esis generating and GWAS validation tools. The studies in this work focus on RNA expression in a popular human model: lymphoblastoid cell lines (LCLs). Chapters 2 and 3 examine microRNA (miRNA) and messenger RNA (mRNA) expression in a total of 300 genotyped human LCLs. The expression of only one miRNA could be associated with a nearby genetic variant. This result was observed in both the African and European samples studied, in a separate val- idation data set, and was technically validated with quantitative PCR. While limited genotype resolution and small sample sizes are likely to be important contributors to this low hit rate, the results strongly suggest experimental con- founders. Highly expressed miRNAs reflected the transformed nature of the cells, highly correlated miRNAs enriched for EBV and malaria associated tar- get mRNA genes, and several miRNAs that were differentially expressed be- tween the European and African samples suggested differential EBV transfer- mation. Chapter 4 presents a study on single cells from some of the same samples, to test the hypothesis that the lack of tissue spatial resolution is an important limiting factor in human genetic epidemiology. Experimental con- founders were also considered: sample growth was found to associate with the expression of several genes. Cell-to-cell gene correlations and distributions made it possible to propose how genes change their expression, functionally differ from each other, and are able to alter their behaviours without altered whole-tissue expression. The results suggest which type of genes are more likely to be susceptible to genetic effects, and propose promoter behaviours altered by genetic variants located near to 13 genes. From these whole-tissue and single cell results the broad conclusion is that, while LCLs are likely to be inappropriate for the study of miRNA genetics, their functional genomics at higher spatial resolution shows promise as a more mechanistic approach for the study of germline genetics.

No description available.