Epigenetics inheritance is a phenomenon where a cell state is inherited across cellular divisions. It is important for maintaining differentiated cells types, and thus is fundamental to the development of multicellular organisms. Failure in the system can lead to deleterious effects that range from embryonic lethality to the development of diseases such as cancer. The Polycomb group (PcG) proteins are important regulators of epigenetic inheritance: they maintain the repression of the Hox genes, which specify the body axes of an organism, throughout development and into adulthood. The loss of function of PcG genes therefore leads to homeotic transformations. A mechanistic understanding of PcG function will yield insight into the mechanisms involved in epigenetics inheritance. The Polycomb Repressive Complex 1 (PRC1) is able to inhibit chromatin remodeling and compact polynucleosomes in vitro; these activities could represent how stable gene repression is achieved in vivo. However, it has been difficult to determine the biological relevance of these activities because the nature of nucleosomal-level compaction in the nucleus is unknown and there is no direct assay for it. To circumvent this problem, we adopted an approach where we introduced into mouse embryonic stem cells (mESCs) and mice the mutations that disrupt the in vitro activities. The mutations are specific lysine-to-alanine and arginine-to-alanine substitutions in a positively charged domain of the CBX2 subunit of PRC1. These are known to disrupt PRC1 polynucleosmal compaction activities from previous studies. We compared the mutant mESCs and mice to wild type counterparts for PcG-related phenotypes. We observed that the mutations prevented the repression of CBX2 (also PRC1) target genes in mESCs. They did not prevent CBX2 from forming PRC1 or binding to chromatin. This suggests that the in vitro activities are specifically involved in inhibiting transcription. Moreover, we observed that the Cbx2 mutant mice exhibit the classic PcG phenotype, which is homeotic transformation in the anterior-posterior axis. The posterior transformations are similar to those observed in the Cbx2-/- mice, indicating that our mutations mimic loss-of-function of Cbx2. Notably, the Cbx2-/- mice had other phenotypes in addition to the PcG phenotype, which we did not observe in our mutants. This indicates that the mutations in the compaction domain of CBX2 specifically disrupted a PcG function during development. These experimental outcomes are consistent with the hypothesis that the in vitro compaction activity of PRC1 is important for how heritable gene repression can be achieved in vivo. / Medical Sciences

Sequencing exomes—the 1% of the genome that codes for proteins—has increased the rate at which the genetic basis of a patient’s disease is determined. Unfortunately, when a patient does not carry a well-established pathogenic variant, it is extremely challenging to establish which of the tens of thousands of variants identified in that individual is contributing to their disease. In these situations, variants must be prioritized to make further investigation more manageable. In this thesis, we have focused on creating statistical frameworks and models to aid in the interpretation of rare variants and towards establishing gene-level metrics for variant prioritization. We developed a sensitive and specific workflow to detect newly arising (de novo) variants from exome sequencing data of parent-child trios, and created a sequence-context based mutational. This mutational model was the basis of a rigorous statistical framework to evaluate the significance of de novo variant burden not only globally, but also per gene. When we applied this framework to de novo variants identified in patients with an autism spectrum disorder, we found a global excess of de novo loss-of-function variants as well as two genes that harbored significantly more de novo loss-of-function variants than expected. We also used the mutational model to predict the expected number of rare (minor allele frequency < 0.1%) variants in exome sequencing datasets of reference individuals. We found a significant depletion of missense and loss-of-function variants in a subset of genes, indicating that these genes are under strong evolutionary constraint. Specifically, we identified 3,230 genes that are intolerant of loss-of-function variation and that set of genes is enriched for established dominant and haploinsufficient disease genes. Similarly, we searched for regions within genes that were intolerant of missense variation. The most missense depleted 15% of the exome contains 83% of reported pathogenic variants found in haploinsufficient disease genes that cause severe disease. Additionally, both gene-level and region-level constraint metrics highlight a set of de novo variants from patients with a neurodevelopmental disorder that are more likely to be pathogenic, supporting the utility of these metrics when interpreting rare variants within the context of disease. / Medical Sciences

The study of life histories is fundamental to understanding why some organisms live for a very short time while others live for a long time, why some produce thousands of offspring while others produce one, or why some need a mate to reproduce while others can do it on their own. The life histories of many animals and plants are well known because we can easily walk into a forest or field and measure them. Fungi, on the other hand, are hard to find and see even though they occur most everywhere in great numbers. In the three chapters of this dissertation, I uncover key aspects of the life history of the genetic model Neurospora crassa, a filamentous fungus. First, I present a novel algorithm for the design of crossing experiments, which test the reproductive abilities of individuals. The algorithm identifies a set of individuals (a “crossing-set”) from a larger pool of potential crossing-sets by maximizing the diversity of traits of interest, for example, maximizing the range of genetic and geographic distances between individuals included in the crossing-set. Second, I use the algorithm to select strains for a mating experiment designed to test for maternal effects—the impact of the mother’s phenotype on the offspring’s phenotype—across sexual reproduction in N. crassa. I measured offspring phenotypes from crosses of all possible pairs of 22 wild-isolated strains. Crosses encompassed reciprocals of 11 mating-type “A” and 11 mating-type “a” strains. After controlling for the genetic and geographic distances between strains in any individual cross, I found strong evidence for maternal control of perithecia (sporocarp) production, as well as maternal effects on spore numbers and spore germination. However, both parents exert equal influence on the percentage of spores that are pigmented, and size of pigmented spores. This experiment is proof of maternal effects in a fungus, an unprecedented discovery. Third, I compared the sexual spore viability data from the maternal effects experiment to asexual spore viability data I gathered for the same strains. I found a striking trade-off between the viabilities of sexual and asexual spores. I used genome-wide polymorphism data to determine the genetic basis of this trade-off and identified 36 single nucleotide polymorphisms (SNPs) significantly associated with both traits. All 36 of the identified SNPs exhibit antagonistic pleiotropy—they have opposing effects on asexual and sexual spore viability. Some of these SNPs occur in well-characterized genes involved in sexual reproduction, asexual reproduction, or both. Using the entire dataset of over 50,000 genome-wide SNPs, I show a negative correlation between effects SNPs on sexual and asexual viability across the entire N. crassa genome. These results suggest that a life history trade-off between asexual and sexual spore viabilities drives a large portion of the genetic variation in the N. crassa genome. / Biology, Organismic and Evolutionary

Abstract: Osteoarthritis (OA) affects 250 million people worldwide. Currently, no targets for disease-modifying osteoarthritis drugs exist. Matrix metalloproteinase-13 (MMP-13) would make it an ideal target; however, its broad biological effects restrict its application as a target enzyme of inhibitory drugs in the treatment of OA. The expression and activation of discoidin domain receptor 2 (DDR2) is increased in human OA tissues and mouse models of OA and was co-localized with elevated expression of MMP-13 in degenerative articular cartilages. In healthy articular cartilage, DDR2 is kept inactivated by the pericellular matrix, which separates the receptor from its ligand, type II collagen. Once enzymes capable of degrading the pericellular molecules expose chondrocytes to type II collagen, DDR2 is activated and induces expression of MMP-13 leading to degradation of type II collagen and proteoglycans resulting in joint destruction and OA. We tested the hypothesis that complete removal of Ddr2 from the knee joint of mouse adult articular cartilage can delay progression of osteoarthritis prior to or after initiation of articular cartilage degeneration. To accomplish this goal, conditional knock out techniques were used with Aggrecan-CreERT2 mice and floxed Ddr2 mice, Ddr2 was removed from articular cartilage of knee joints in mice at 8 weeks of age via intraperitoneal Tamoxifen injection (2mg/10g body weight) for 5 consecutive days (Group A). Mice were subjected to destabilization of the medial meniscus (DMM) or sham surgery at 10 weeks of age. An additional experimental group was subjected to DMM or sham surgery at 10 weeks of age and then DDR2 was removed by intraperitoneal Tamoxifen injection 8 weeks later (Group B). Knee joints from mice in Group A and their corresponding controls were harvested at 8 weeks or 16 weeks post-surgery and mice from Group B and their controls were harvested at 16 weeks post surgery. Histology was performed and the OARSI Modified Mankin Score was used to evaluate articular cartilage degeneration. Statistically significant differences were determined via T-test. We found the average modified score for Group A 8 week control was 1.64 (n=7) whereas with Ddr2 removed was 0.64 (n=7) [P<0.05]. 2) The average modified score for Group A 16 week control was 4.67 (n=7) and with Ddr2 removed was 1.27 (n=9) [P<0.05]. 3) The average modified score for Group B was 1.1 (n=5). In conclusion, conditional removal of Ddr2 in articular cartilage attenuated articular cartilage degeneration in mature knee joints of mouse models of OA.

<p> There are approximately 230,000 new diagnoses of prostate cancer every year in the U.S., making prostate cancer the most diagnosed cancer in men. It is responsible for approximately 30,000 deaths per year, with only lung cancer taking more lives. An important distinction must be made in men with prostate cancer. The majority of men with prostate cancer have a relatively indolent form of the disease, meaning high survival rates (100% survival 5 years after diagnosis) and no invasion of the tumor to other organs. However, approximately 4% of men are diagnosed with an aggressive form of the disease, and for these men, the survival rate is a mere 30% after 5 years. And for many patients, it is clinically difficult to differentiate between the indolent and the aggressive forms of the disease. Therefore, it is imperative to develop new genetic models of prostate cancer, and the mouse is an excellent model organism in which to do so. In 2009, mice were used to discover a new type of stem cell, called a castration-resistant Nkx3.1-?expressing cell in the luminal cell population of the prostate. We have used a mouse model targeting these cells to study the roles of two tumor suppressors, adenomatous polyposis coli (Apc) and Smad4. Apc down-regulates the Wnt signaling pathway, which is a carcinogenic pathway in the prostates of humans and mice. Deletion of Apc in mice causes an increase of Wnt signaling and prostate cells to proliferate but not invade, which represents a relatively indolent, precancerous phenotype. Smad4 is a transcription factor that controls the signaling of two pathways: transforming growth factor &beta; and bone morphogenetic protein signaling. Deletion of Smad4 causes these pathways to shut off. When Apc and Smad4 are deleted simultaneously, mice develop aggressive, invasive prostate cancer. This work suggests that these two tumor suppressors &ndash; and the pathways they control &ndash; are important regulators of prostate cancer, could allow for clinicians to differentiate between indolent and aggressive disease, and should be targeted therapeutically in prostate cancer patients.</p><p>

<p> The word "cancer" rarely stands alone, usually prefaced with its anatomical location: lung cancer, prostate cancer, brain cancer. With the advancement of high-throughput omics approaches, specific oncogenic events are reorganizing the landscape of cancer classification, at once creating commonalities between cancers arising in diverse anatomical locations and dividing organ-centric classifications of cancer into a multitude of subtypes. The term "precision medicine" postulates that these new, data-driven groupings based on molecular characterization are the key to making rational therapeutic choices. </p><p> The majority of this dissertation addresses the disconnect between extensive molecular characterization and poor cancer therapy outcomes for patients with glioblastoma multiforme (GBM). Despite clear evidence that hyperproduction of the ligand for PDGFR (platelet-derived growth factor receptor &alpha;) is sufficient to generate GBM of the proneural subtype, anti-PDGFR&alpha; therapeutics have proven disappointing in clinical trials. Cell adaptation contributes to therapeutic escape. In GBM, proneural tumor cells adopt transcriptional profiles of the mesenchymal subtype. The interconversion between the proneural and mesenchymal transcriptional classes within a tumor population presents both a challenge and an opportunity for therapeutic approaches. The proneural subtype has a proliferation phenotype and presents druggable targets such as PDGFR&alpha;. The mesenchymal subtype presents an invasive phenotype, but the targets are more challenging to drug. The typical screening for combination therapies that synergize to induce cell death is not as advantageous here, where the disease management is expected to include cytostatic drugs that act on two different aspects of the phenotype: proneurally mediated proliferation and mesenchymally mediated invasion. This work examines the applicability of a combination approach against a proneural target, PDGFR&alpha;, and mesenchymal targets in the STAT3 (signal transducer and activator of transcription 3) pathway, in the context of a proneural model of GBM.</p><p> The work is concluded with collection of work applying precision medicine in other disease contexts, most notably canine lymphoma.</p><p>

With the completion of the sequencing of the human genome near, the next phase will be to identify and annotate all transcriptional units. We have utilized a procedure that directly selects cDNA from genomic DNA in order to isolate putative transcriptional units for identification of novel genes at the 4q34 locus. In this procedure, cDNA fragments were isolated following the hybridization of cDNA pools to 7 BAC clones spanning the 4q34 region. The 4q34 region is approximately 2.0 Mb and contains 7 known genes and 8 predicted genes. In addition, EST evidence annotated in the public databases supports the notion that there are additional transcriptional units in this region. The primary cDNA pool used in this procedure was generated with a universal primer for amplification and cloning. Approximately 350 clones were analyzed by automated fluorescence sequencing and 26 clones were shown to originate from DNA at the 4q34 locus. The other clones included rRNA, mitochondrial DNA, repetitive sequences and low-copy repeat elements, similar to the results obtained in comparable cDNA selection attempts. The tentative transcriptional units have been arranged on the current map of the 4q34 region. This map will provide insight into the organization and function of this chromosome, and provide the preliminary framework for a detailed transcription map of the region. Furthermore, novel gene identification at this locus will provide candidates for Parkinson's disease, which has been mapped to this region by our laboratory.

To facilitate detection of genotoxicity from environmental mutagen exposure we generated an in vitro green fluorescence protein (GFP) activation assay that quickly and effectively detects frameshift mutations in tandem repeat sequences (TRS). Two murine cell lines were stably transfected with GFP reporter plasmids in which the GFP constructs contain TRS that shift the GFP coding sequence out of frame. These included several 2-6bp repeat sequences, a control non-repetitive sequence and a human gene sequence with TRS. Transfected cultures were exposed to five model chemicals: hydrogen peroxide (H2O2), 12-O-tetradecanoyl-phorbol-13-acetate (TPA), benzo-a-pyrene-diol-epoxide (BPDE), ethyl nitrosourea (ENU), 9-aminoacridine (9AA). Frameshift mutations resulted in green fluorescent revertants, as determined by flow cytometry and confirmed by sequencing. All five treatments induced a statistically significant sequence- and dose-dependent response in both cell lines. Results from these experiments reveal that the assay responds robustly to various classes of mutagenic substances, as well as carcinogens that are inactive in conventional mutation assays.

Vertebrate Dlx genes code for homeodomain (a conserved DNA binding domain) transcription factors involved in the development of the forebrain, craniofacial skeleton, dentition, sensory organs and limbs. Vertebrates have at least 6 Dlx genes whereas invertebrates have only one Dlx homologue (called the distal-less or Dll gene) (Stock, Ellies et al. 1996). Vertebrate Dlx genes are organized as pairs of convergently transcribed genes. The intergenic region between each of these bigenes contains at least two highly conserved sequences (300-500bp each) that act as enhancers of transcription of the flanking Dlx genes. Activity of these enhancers when tested with reporter genes (GFP and LacZ) in transgenic zebrafish and mice recapitulates the endogenous Dlx expression patterns. Here we are investigating Dlx regulation in lamprey, an agnathan species and a sister group of gnathostomes. In doing this we hope to (1) trace back the origin of the conserved Dlx intergenic enhancers found in vertebrates and (2) address the question, can the differences in regulation of lamprey Dlx genes with that of gnathostomes account for some of the morphological differences seen in lampreys (like absence of jaws)? We have shown that lamprey have at least four Dlx genes (A, B, C, D), two of which are arranged in the bigenic configuration seen in other vertebrates. We were unable to find the presence of the conserved intergenic enhancers by Southern hybridization and PCR experiments within this lamprey Dlx intergenic region. We have shown that this region can target expression of reporter genes in the forebrain, branchial arches and trunk somites when tested in zebrafish embryos in a reporter gene assay, suggesting many of the Dlx regulatory mechanisms are conserved between lamprey and gnathostomes.

Brachydactyly type A1 (BDA1) is a congenital disorder that affects normal bone development and patterning. Affected individuals have short fingers, broad hands and are generally short in stature. BDA1 was the first human trait described in terms of autosomal dominant Mendelian inheritance. Missense mutations in the Indian hedgehog (IHH) gene have been shown to be responsible for the disorder, and a second locus has been mapped to a region on chromosome 5p13.3-p13.2 in a single kindred. We have since obtained DNA from four new BDA1-affected kindreds. We were successful in identifying IHH mutations in 3 of these families, thus further supporting a role for IHH in the pathogenesis of BDA1. We have also obtained DNA from a consanguineous BDA1-affected kindred in which mutations in IHH, as well as linkage to the chromosome 5p13.3-p13.2 region, have been excluded. Taken together, this data supports existence of a third locus for BDA1. A genome-wide linkage screen was initiated in the family, and linkage was observed at a region on chromosome 20q11.22-q11.23. Subsequent screening of the candidate gene GDF5, located in the chromosome 20 disease locus, revealed a homozygous c.1138C>T nucleotide change in all affected members of the sibship. This mutation is hypothesized to disrupt the highly-conserved tertiary structure of the GDF5 monomer which is critical for the dimerization and maturation of GDF5, thus leading to interrupted GDF5 signaling and severe BDA1 pathogenesis in homozygous individuals. This thesis identifies GDF5 as the second disease gene for BDA1, and describes the first case of BDA1 in a consanguineous family.