Relational Web tables have become an important resource for applications such as factual search and entity augmentation. A major challenge for an automatic identification of relevant tables on the Web is the fact that many of these tables have missing or non-informative column labels. Research has focused largely on recovering the meaning of columns by inferring class labels from the instances using external knowledge bases. The table context, which often contains additional information on the table's content, is frequently considered as an indicator for the general content of a table, but not as a source for column-specific details. In this paper, we propose a novel approach to identify and extract column-specific information from the context of Web tables. In our extraction framework, we consider different techniques to extract directly as well as indirectly related phrases. We perform a number of experiments on Web tables extracted from Wikipedia. The results show that column-specific information extracted using our simple heuristic significantly boost precision and recall for table and column search.

Synaptic plasticity—the ability of a synapse to change—is fundamental to basic brain function and behavioral adaptation. Studying the mechanisms of synaptic plasticity benefits our understanding of the formation of neuronal connections and circuitry, which has great implications in the field of learning and memory and the studies of numerous human diseases. The Drosophila larval neuromuscular junction (NMJ) system is a powerful system for studying synaptic plasticity. The NMJ consists of at least two different types of motorneurons innervating the body wall muscles. Type I motorneurons controls muscle contraction using glutamate as the neurotransmitter, while type II are modulatory neurons that contain octopamine. Octopamine is a potent modulator of behavior in invertebrates. Nevertheless, its function at the synapse is poorly understood. In my thesis research, I investigated the role of octopamine in synaptic plasticity using the Drosophila NMJ system. Preliminary observations indicate that increased larval locomotion during starvation results in an increase of filopodia-like structures at type II terminals. These structures, which we termed as “synaptopods” in our previous studies, contain synaptic proteins and can mature into type II synapses. I demonstrated that this outgrowth of type II terminals is dependent on activity and octopamine. Mutations and genetic manipulations affecting the production of octopamine decrease synaptopods, whereas increase of type II activity or exogenous application of octopamine increase synaptopods. Interestingly, I found that the type II octopaminergic neurons have an absolute dependence on activity for their innervation of the muscles. Blocking activity in these neurons throughout development results in no type II synapses at the NMJ, whereas blocking activity after the formation of synapses results in gradual degradation of type II terminals. Next, I examined the autoregulatory mechanism underlying the octopamine-induced synaptic growth in octopaminergic neurons. I discovered that this positive-feedback mechanism depends on an octopamine autoreceptor, Octß2R. This receptor in turn activates a cAMP- and CREB-dependent pathway that is required in the octopamine-induction of synaptopods. Furthermore, I demonstrated that this octopaminergic autoregulatory mechanism is necessary for the larva to properly increase its locomotor activity during starvation. Thirdly, I investigated the possibility that type II innervation might regulate type I synaptic growth through octopamine. We found that ablation, blocking of type II activity, or the absence of octopamine results in reduced type I outgrowth, and this paracrine signaling is mediated by Octß2R which is also present in type I motorneurons. Lastly, the function of another octopamine receptor, Octß1R, was examined. In contrast to Octß2R, Octß1R is inhibitory to synaptic growth. I demonstrated that the inhibitory effect of this receptor is likely accomplished through the inhibitory G-protein Goα. Similar to Octß2R, Octß1R also regulates the synaptic growth of both type I and type II motorneurons in a cell-autonomous manner. The inhibitory function of this receptor potentially breaks the positive feedback loop mediated by Octß2R, allowing the animal to reset its neurons when the environment is favorable. In summary, the research presented in this thesis has unraveled both autoregulatory and paracrine mechanisms in which octopamine modulates synaptic and behavior plasticity through excitatory and inhibitory receptors.

The Hedgehog (Hh) family of proteins are secreted growth factors that play an essential role in the embryonic development of all organisms and the main components in the pathway are conserved from insects to humans. These proteins affect patterning and morphogenesis of multiple tissues. Therefore, mutations in the Hh pathway can result in a wide range of developmental defects and oncogenic diseases. Because the main components in the pathway are conserved from insects to humans, Drosophilahas been shown to provide a genetically tractable system to gain insight into the processes that Hh is involved in. In this study, the roles of Hh cholesterol modification and endocytosis during gradient fonnation are explored in the Drosophila larval wing imaginal disc. To exclude the possibility of looking at a redistribution of preexisting Hh instead of Hh movement, a spatially and temporally regulated system has been developed to induce Hh expression. Functional Hh-GFP with and without the cholesterol-modification was expressed in a wild-type or shi-tslendocytosis mutant background. The Gal80 system was used to temporally express (pulse) the Hh-GFP transgenes to look at the rate of Hh gradient formation over time and determine whether this process was affected by cholesterol modification and/or endocytosis. Hh with and without cholesterol were both largely detected in punctate structures and the spreading of the different forms of Hh was quantified by measuring distances of these particles from the expressing cells. Hh without cholesterol showed a greater range of distribution, but a lower percentage of particles near the source. Loss of endocytosis blocked formation of intracellular Hh particles, but did not dramatically alter its movement to target cells. Staining for Hh, its receptor Ptc and cortical actin revealed that these punctate structures could be classified into four types of Hh containing particles: cytoplasmic with and without Ptc, and cell surface with and without Ptc. Cholesterol is specifically required for the formation of cytoplasmic particles lacking Ptc. While previous studies have shown discrepancies in the localization of Hh following a block in endocytosis, Hh with and without cholesterol is detected at both apical and basolateral surfaces, but not at basal surfaces. In the absence of cholesterol and endocytosis, Hh particles can be observed in the extracellular space. Through three-dimensional reconstruction and quantitative analysis, this study concludes that the cholesterol modification is required to restrict Hh movement. In addition, the cholesterol modification promotes Ptc-independent internalization. This study also observes that Dynamin-dependent endocytosis is necessary for internalization but does not play an essential role in Hh distribution. The data in this thesis supports the model in which Hh movement occurs via planar diffusion.

Since the discovery in 1993 of the first small silencing RNA, a dizzying number of small RNAs have been identified, including microRNAs (miRNAs), small interfering RNAs (siRNAs) and Piwi-interacting RNAs (piRNAs). These classes differ in their biogenesis, modes of target regulation and in the biological pathways they regulate. Historically, siRNAs were believed to arise only from exogenous double-stranded RNA triggers in organisms lacking RNA-dependent RNA polymerases. However, the discovery of endogenous siRNAs in flies expanded the biological significance of siRNAs beyond viral defense. By high throughput sequencing we identified Drosophila endosiRNAs as 21 nt small RNAs, bearing a 2´-O-methyl group at their 3´ ends, and depleted in dicer-2 mutants. Methylation of small RNAs at the 3´ end in the soma, is a consequence of assembly into a mature Argonaute2-RNA induced silencing complex. In addition to endo-siRNAs, we observed certain miRNAs or their miRNA* partners loading into Argonaute2. We discovered, that irrespective of its biogenesis, a miRNA duplex can load into either Argonaute (Ago1 or Ago2), contingent on its structural and sequence features, followed by assignment of one of the strands in the duplex as the functional or guide strand. Usually the miRNA strand is selected as the guide in complex with Ago1 and miRNA* strand with Ago2. In our efforts towards finding 3´ modified small RNAs in the fly soma, we also discovered 24-28nt small RNAs in certain fly genotypes, particularly ago2 and dcr-2mutants. 24-28nt small RNAs share many features with piRNAs present in the germline, and a significant fraction of the 24-28nt small RNAs originate from similar transposon clusters as somatic endo-siRNAs. Therefore the same RNA can potentially act as a precursor for both endo-siRNA and piRNA-like small RNA biogenesis. We are analyzing the genomic regions that spawn somatic small RNAs in order to understand the triggers for their production. Ultimately, we want to attain insight into the underlying complexity that interconnects these small RNA pathways. Dysregulation of small RNAs leads to defects in germline development, organogenesis, cell growth and differentiation. This thesis research provides vital insight into the network of interactions that fine-tune the small RNA pathways. Understanding the flow of information between the small RNA pathways, a great deal of which has been revealed only in the recent years, will help us comprehend how the pathways compete and collaborate with each other, enabling each other’s optimum function.

Understanding the mechanisms of memory is vital in making sense of the continuity of the self, our experience of time and of the relation between mind and body. The invertebrate Drosophila melanogaster offers us an opportunity to study and comprehend the overwhelming complexity of memory on a smaller scale. The work presented here investigates the neural circuitry in the fly brain required for olfactory memory processing. Our observation that Dorsal Paired Medial (DPM) neurons, which project only to mushroom body (MB) neurons, are required during memory storage but not for acquisition or retrieval, led us to revisit the role of MB neurons in memory processing. We show that neurotransmission from the α'β' subset of MB neurons is required to acquire and stabilize aversive and appetitive odor memory but is dispensable during memory retrieval. In contrast neurotransmission from MB αβ neurons is only required for memory retrieval. These data suggest a dynamic requirement for the different subsets of MB neurons in memory and are consistent with the notion that recurrent activity in a MB α'β' neuron-DPM neuron loop is required to consolidate memories formed in the MB αβ neurons. Furthermore, we show that a single two-minute training session pairing odor with an ethologically relevant sugar reinforcement forms long-term appetitive memory that lasts for days. This robust, stable LTM is protein-synthesis-, Creb- and radish-dependent and relies on the activity in the DPM neuron and mushroom body α'β' neuron circuit during the first hour after training and mushroom body αβ neuron output during retrieval. Lastly, experiments feeding and/or starving flies after training reveals a critical motivational drive that enables memory retrieval. Neural correlates of motivational states are poorly understood, but using our assay we found a neural mechanism that accounts for this motivation-state-dependence. We demonstrate a role for the Neuropeptide F (dNPF) circuitry, which led to the identification of six dopaminergic MB-MP neurons that innervate the mushroom bodies as being critical for appetitive memory performance. Directly blocking the MB-MP neurons releases memory performance in fed flies whereas stimulating them suppresses memory performance in hungry flies. These studies provide us with an enhanced knowledge of systems level memory processing in Drosophila.

Post-transcriptional regulation of gene expression is a mechanism that governs developmental and cellular events in metazoans. In early embryogenesis, transcriptionally quiescent cells depend upon maternally supplied factors such as RNA binding proteins and RNA that control key decisions. Morphogen gradients form and in turn pattern the early embryo generating different cell types and spatial order. In the nematode Caenorhabditis elegans, the early embryo relies upon several RNA binding proteins that control mRNA stability, translation efficiency, and/or mRNA localization of cell fate determinants essential for proper development. MEX-5 and MEX-3 are two conserved RNA-binding proteins required to pattern the anterior/posterior axis and early embryo. Mutation of either gene results in a maternal effect lethal phenotype with proliferating posterior muscle into the anterior blastomeres (Muscle EXcess). Several cell-fate determinants are aberrantly expressed in mex-5 and mex-3 embryos. Both proteins are thought to interact with cis-regulatory elements present in 3’-UTRs of target RNAs controlling their metabolism. However, previous studies failed to demonstrate that these proteins regulate maternal transcripts directly. This dissertation presents a thorough assessment of the RNA binding properties of MEX-5 and MEX-3. Quantitative biochemical approaches were used to determine the RNA binding specificity of both proteins. MEX-5 has a relaxed specificity, binding with high affinity to linear RNA containing a tract of six or more uridines within an eight-nucleotide window. This is very different from its mammalian homologs Tristetraprolin (TTP) and ERF-2. I was able to identify two amino acids present within the MEX-5 RNA binding domain that are required for the differential RNA recognition observed between MEX-5 and TTP. MEX-3 on the other hand is a specific RNA binding protein, recognizing a bipartite element with flexible spacing between two four-nucleotide half-sites. I demonstrate that this element is required for MEX-3 dependent regulation in vivo. Previous studies only identify a small number of candidate regulatory targets of MEX-5 and MEX-3. The defined sequence specificity of both proteins is used to predict new putative targets that may be regulated by either protein. Collectively, this study examines the RNA binding properties of MEX-5 and MEX-3 to clarify their role as post-transcriptional regulators in nematode development.

Upon fertilization, the early embryo sustains most of the cellular processes using the maternally deposited reserves in the egg itself until the zygotic gene expression takes charge. Among the plethora of essential components provided by the mother are small non-coding RNAs called PIWI-interacting RNAs (piRNAs), which provide immunity to the zygote against transposon challenge. In this thesis, I have presented three different functions of piRNAs in Drosophila melanogaster- in maintenance of genomic integrity, telomere protection and their role as an adaptive immune system against genomic parasites. In Chapter 2, I have described the phenotypic effects of the loss of piRNA function in early embryos. The mutations affecting the piRNA pathway are known to cause embryonic lethality. To describe this lethality in detail, I have shown that all the characterized piRNA mutants show compromised zygotic genomic integrity during early embryogenesis. In addition, two piRNA pathway components, Aubergine (Aub) and Armitage (Armi) are also required for telomere resolution during early embryogenesis. Aub and Armi recruit telomeric protection complex proteins, HOAP and HP1, to the telomeric ends and thus avoid activation of the Non-homologous end joining (NHEJ) DNA repair pathway at the telomeres. There are about 120 transposon families in Drosophila melanogaster and piRNA pathway mutations cause activation of many of the resident transposons in the genome. In Chapter 3, I have described the effects of infection by a single transposon, P-element, in naïve strains by introduction through the zygote. Activation of the P-element leads to desilencing of unrelated transposons, causing accumulation of germline DNA damage which is linked to severely reduced fertility in the hybrid females. However, there is partial restoration of fertility as the hybrid progeny age, which correlates with P-element piRNA production and thus P-element silencing. Additionally, a number of transposons mobilize into piRNA generating heterochromatic clusters in the genome, and these insertions are stably inherited in the progeny. Collectively our data shows that piRNA production can be triggered in the adults in an absence of maternal contribution and that piRNAs serve as an adaptive immune system which helps resolve an internal genetic conflict between the host and the parasite. In an effort to understand the phenotypic effects of piRNA dysfunction in Drosophila, we have uncovered new exciting roles for piRNAs in development and presented evidence how transposons can act as architects in restructuring the host genome.

The nervous system is made up of two major cell types, neurons and glia. The major distinguishing feature between neuronal cells and glial cells is that neurons are capable of transmitting action potentials while glial cells are electrically incompetent. For over a century glial cells were neglected and it was thought they existed merely to provide trophic and structural support to neurons. However, in the past few decades it has become increasingly clear that glial cell functions underlie almost all aspects of nervous system development, maintenance, and health. During development, glia act as permissive substrates for axons, provide guidance cues, regulate axon bundling, facilitate synapse formation, refine synaptic connections, and promote neuronal survival. In the mature nervous system glial cells regulate adult neurogenesis through phagocytosis, act as the primary immune cell, and contribute to complex processes such as learning and memory. In recent years, glial cells have also become a primary focus in the study of neurodegenerative diseases. Mounting evidence shows that glial cells exert both beneficial as well as detrimental effects in the pathology of several nervous system disorders, and modulation of glial activity is emerging as a viable therapeutic strategy for many diseases. Although glial cells are critical to the proper development and functioning of the nervous system, there is still relatively little known about the molecular mechanisms used by glial cells, how they exert their effects on neurons, and how glia and neurons communicate. Despite the relative simplicity and small size of the Drosophila nervous system, glial cell organization and function in flies shows a remarkable complexity similar to vertebrate glial cells. In this study I use Drosophila as a model organism to study cellular and molecular mechanisms of glial clearance of axonal debris after acute axotomy. In chapter two of this thesis, I characterize three distinct subtypes of glial cells in the adult brain; cell body glia which ensheath neuronal cell bodies in the cortex region of the brain, astrocyte like glial cells which bear striking morphological similarity to mammalian astrocytes and share common molecular components, and ensheathing glial cells which I show act as the primary phagocytic cell type in the neuropil region of the brain. In addition, I identify dCed-6, the ortholog of mammalian GULP, as a necessary component of the glial phagocytic machinery. In chapter three of this thesis, I perform a candidate based, in vivo, RNAi screen to identify novel genes involved in the glial engulfment of degenerating axon material. The Gal4/UAS system was used to drive UAS-RNAi for approximately 300 candidate genes with the glial specific repo-Gal4 driver. Two assays were used as a readout in this screen, clearance of axon material five days after injury, and Draper upregulation one day after maxillary palp or antennal injury. Overall, I identified 20 genes which, when knocked down specifically in glial cells, result in axon clearance defects after injury. Finally, in chapter four I identify Stat92E as a novel glial gene required for glial phagocytic function. I show that Stat92E regulates both basal and injury induced Draper expression. Injury-induced Draper expression is transcriptionally regulated through a Stat92E dependent non-canonical signaling mechanism whereby signaling through the Draper receptor activates Stat92E which in turn transcriptionally activates draper through a binding site located in the first intron of Draper. Draper represents only the second receptor known to positively regulate Stat92E transcriptional activity under normal physiological conditions.

Costimulation blockade based on a donor-specific transfusion and anti-CD154 mAb is effective for establishing mixed allogeneic hematopoietic chimerism and inducing transplantation tolerance. Despite its potential, recent evidence suggests that the efficacy of costimulation blockade can be reduced by environmental perturbations such as infection or inflammation that activate toll-like receptors (TLR). TLR agonists prevent costimulation blockade-induced prolongation of solid organ allografts, but their effect on the establishment of hematopoietic chimerism has not been reported. In this dissertation, we hypothesized that TLR activation during costimulation blockade would prevent the establishment of mixed hematopoietic chimerism and shorten skin allograft survival. To test this hypothesis, costimulation blockade-treated mice were co-injected with TLR2 (Pam3Cys), TLR3 (poly I:C), or TLR4 (LPS) agonists and transplanted with allogeneic bone marrow and skin grafts. Supporting our hypothesis, we observed that TLR agonists administered at the time of costimulation blockade prevented the establishment of mixed hematopoietic chimerism and shortened skin allograft survival. To investigate underlying cellular and molecular mechanisms, we first determined that LPS administration during costimulation blockade did not increase production of alloantibodies or activate natural killer cells. Similarly, costimulation blockade-treated mice depleted of CD4+ or CD8+ cells did not become chimeric when co-injected with LPS. In contrast, mice depleted of both CD4+ and CD8+cell subsets were resistant to the effects of LPS. We next observed that alloreactive T cells were activated by TLR agonists in mice treated with costimulation blockade, and this activation correlated with LPS-induced maturation of donor and host alloantigen-presenting cells. In contrast, TLR4-deficient mice treated with costimulation blockade and LPS did not upregulate costimulatory molecules on their APCs, and mixed chimerism and permanent skin allograft survival were readily achieved. We further observed that injection of recombinant IFN-β recapitulated the detrimental effects of LPS, and that LPS-injected mice deficient in the type I IFN receptor were partially protected. Importantly, alloantigen-presenting cells did not upregulate costimulatory molecules in response to LPS, and mixed chimerism and permanent skin allograft survival were readily established in type I IFN receptor and MyD88 double deficient mice treated with costimulation blockade. We conclude that the TLR4 agonist LPS prevents the establishment of mixed hematopoietic chimerism and shortens skin allograft survival in mice treated with costimulation blockade by inducing the production of type 1 IFN and MyD88-dependent factors that upregulate costimulatory molecules on APCs, leading to the generation of activated alloreactive T cells.

From the yeast genome completed in 1996 to the 12 Drosophilagenomes published earlier this year; little more than a decade has provided an incredible amount of genomic data. Yet even with this mountain of genetic information the regulatory networks that control gene expression remain relatively undefined. In part, this is due to the enormous amount of non-coding DNA, over 98% of the human genome, which needs to be made sense of. It is also due to the large number of transcription factors, potentially 2,000 such factors in the human genome, which may contribute to any given network directly or indirectly. Certainly, one of the central limitations has been the paucity of transcription factor (TF) specificity data that would aid in the prediction of regulatory targets throughout a genome. The general lack of specificity data has hindered the prediction of regulatory targets for individual TFs as well as groups of factors that function within a common regulatory pathway. A large collection of factor specificities would allow for the combinatorial prediction of regulatory targets that considers all factors actively expressed in a given cell, under a given condition. Herein we describe substantial improvements to a previous bacterial one-hybrid system with increased sensitivity and dynamic range that make it amenable for the high-throughput analysis of sequence-specific TFs. Currently we have characterized 108 (14.3%) of the predicted TFs in Drosophilathat fall into a broad range of DNA-binding domain families, demonstrating the feasibility of characterizing a large number of TFs using this technology. To fully exploit our large database of binding specificities, we have created a GBrowse-based search tool that allows an end-user to examine the overrepresentation of binding sites for any number of individual factors as well as combinations of these factors in up to six Drosophila genomes (veda.cs.uiuc.edu/cgi-bin/gbrowse/gbrowse/Dmel4). We have used this tool to demonstrate that a collection of factor specificities within a common pathway will successfully predict previously validated cis-regulatory modules within a genome. Furthermore, within our database we provide a complete catalog of DNA-binding specificities for all 84 homeodomains in Drosophila. This catalog enabled us to propose and test a detailed set of recognition rules for homeodomains and use this information to predict the specificities of the majority of homeodomains in the human genome.