Childhood blindness due to hereditary retinal disease represents a huge clinical problem. Morbidity, blindness but also a major lack of understanding the molecular genetic events that lead to blindness in very young children hamper clinical management of this important and common problem. One way to get entry into the disease mechanism of childhood retinal blindness is to study Mendelian disease entities that are caused by relatively straightforward links between mutations in genes that lead to aberrant retinal protein function and the resulting disease phenotype. These discoveries in the Mendelian model would lead to understanding of the genetics, proteomics and most importantly retinal disease mechanisms and cycles that underlie the more complex diseases causing blindness. We have assembled a very exciting collection of 21 consanguineous pedigrees with a devastating autosomal recessive Mendelian disease, nonsyndromic congenital retinal nonattachment (NCRNA) and 16 families with autosomal recessive retinitis pigmentosa (arRP). Our main guiding hypothesis is that a significant portion of NCRNA will be caused by mutations in novel retinal genes as currently only one NCRNA gene is known (ATOH7). These novel genes may give us insight into the disease pathways of retinal detachment itself, a huge and clinically unmet medical need. Adult retinal detachments are common and the disease pathways are unknown.In a preliminary study of our cohort, we performed SNP genotyping where we aimed to identify homozygous (hmz) regions. Homozygous regions contain the causal gene(s) in consanguineous families. We found that in many of our patients known retinal detachment or RP gene(s) resided in one of the top 5 largest homozygous segments documented by the SNP array. Direct sequencing of these genes identified the causal mutations in known retinal genes in 11 out of our 21 retinal detachment families and 2 out of our 16 RP families. Thus, 35% of the patients were successfully genotyped in our study with six novel and seven previously reported mutations in known genes. Interestingly, 65% of our cohort of arRP and NCRNA patients remains genetically undetermined, despite our intensive search. After our prescreening for known mutations in NCRNA and RP genes using a combination of conventional genotyping methods, the genetic defects in 10 NCRNA and 14 RP patients remained unexplained suggesting new gene(s) in those patients. Although the clinical diagnosis of all children was NCRNA, in eight families, the molecular diagnosis determined that the disease process was in fact, familial exudative vitreoretinopathy (FEVR), a rare inherited vitreoretinal dystrophy characterized by the disruption of retinal vascular development (FEVR; MIM 133780), caused by pathogenic variation occurred in LRP5, TSPAN12 and NDP genes. However, the phenotype remains the very severe congenital retinal detachment, previously not known to be associated with FEVR mutations. Therefore, we have expanded the phenotypic spectrum of FEVR, a severe retinal detachment phenotype that clinically overlaps with NCRNA. Also, we identified in our Pakistani cohort, the previously identified large deletion (6523 bp del) in ATOH7 found in the Kurdish founder population of Northern Iranian with a high incidence of NCRNA (Ghiasvand et. al, 2011) [5]. We therefore establish for the first time genetic overlap between the Iranian and Pakistani populations. / La cécité infantile causée par des maladies héréditaires de la rétine représente un énorme problème clinique. La morbidité, la cécité mais aussi le manque important de la compréhension des événements moléculaires génétiques conduisant à la cécité chez les très jeunes enfants diminuent la capacité de gérer cliniquement ce problème important et commun. Un moyen de comprendre le mécanisme de la maladie de la cécité infantile est d'étudier les maladies mendéliennes causées par des liens relativement simples entre les mutations génétiques, qui rendent aberrant le fonctionnement des protéines rétiniennes, et le phénotype de la maladie qui en résulte. Ces découvertes dans le modèle mendélien conduiraient à la compréhension des mécanismes des maladies rétiniennes ainsi que d'autres maladies plus complexes provoquant la cécité. Nous avons rassemblé une collection très intéressante de 21 pedigrees consanguins dont les membres possèdent une maladie autosomique récessif aux effets dévastatrices: nonsyndromic congenital retinal nonattachement (NCRNA) ainsi que 16 familles avec rétinite pigmentaire autosomique récessive (arRP). Notre principale hypothèse de base est que une partie importante de NCRNA sera causée par des mutations de nouveaux gènes rétiniennes puisque actuellement un seul gène NCRNA est connu (ATOH7). Ces nouveaux gènes peuvent nous donner des informations sur la mécanisme de la maladie de la détachement rétinienne. Décollement de la rétine chez les adultes sont fréquents et les voies de la maladie sont inconnues. Dans une étude préliminaire de notre cohorte, nous avons effectué de génotypage de SNPs où nous avons cherché à identifier des régions homozygotes (hmz). Les régions homozygotes contiennent le gène responsable dans des familles consanguines. Nous avons constaté que, parmi beaucoup de nos patients, les gènes de la détachement rétinienne ou de RP résidaient dans l'un des 5 plus grands secteurs homozygotes documentés par l'ensemble SNP. Le séquençage direct de ces gènes nous a permis d'identifier les mutations responsables dans les gènes rétiniens connus pour 11 sur nos 21 familles de décollement de la rétine et 2 sur nos 16 familles de RP. Conséquemment, six nouvelles mutations et sept mutations rapportées dans des gènes connus se trouvent dans 35% des patients. Fait intéressant, 65% de notre cohorte de patients arRP et NCRNA demeurent génétiquement indéterminée, en dépit de notre recherche intensive. Après notre présélection pour les mutations connues dans les gènes NCRNA et RP en utilisant une combinaison de méthodes de génotypage conventionnelles, les défauts génétiques dans 10 patients NCRNA et 14 patients RP sont restées inexpliqués suggérant de nouveaux gène(s) chez ces patients. Bien que le diagnostic clinique de tous les enfants était NCRNA, dans huit familles, le diagnostic moléculaire a déterminé que le processus de la maladie était en fait familial exudative vitreoretinopathy FEVR, une rare dystrophie vitréo-rétinienne héréditaire caractérisée par la perturbation du développement vasculaire rétinienne (FEVR; MIM 133780), causé par des variations pathologiques survenues dans les gènes LRP5, TSPAN12 et NDP. Cependant, le phénotype reste d'être des détachements de la rétine congénitale sévères qui n'étaient pas connue pour être associée à des mutations FEVR. Par conséquent, nous avons élargi le spectre phénotypique de la FEVR, un phénotype de détachement rétinien sévère contenant des caractéristiques cliniques similaires à NCRNA. En outre, Nous avons trouvé, dans notre cohorte pakistanaise, la suppression de grande taille (6523 pb del) dans ATOH7 préalablement identifiée parmi la population fondatrice kurde du nord de l'Iran chez qui on retrouve une incidence élevée de NCRNA (Ghiasvand et. al., 2011). Nous avons donc établi, pour la première fois, le chevauchement génétique entre les populations iraniennes et pakistanaises.

In 2005, Simons et al. proposed that during kidney development there is a switch from canonical to non-canonical WNT signaling. Failure to do so could result in a cystic kidney phenotype. Our lab described that canonical WNT signaling is high during kidney development and is down regulated once nephrogenesis is completed in both wild type and cystic mice. We hypothesize that non-canonical WNT signaling could be abnormally regulated in autosomal dominant polycystic kidney disease (ADPKD). We studied the role of the WNT-Ca2+ non-canonical pathway during kidney development and explored a potential role in ADPKD. We employed an NFAT-luciferase (NFAT-Luc) reporter mouse, where NFAT is the primary downstream signaling molecule of the WNT-Ca2+ pathway. This reporter was responsive to non-canonical WNT5a and WNT11, and the effect of the WNTs could be abrogated by cyclosporin-A, a specific inhibitor of the pathway. In situ hybridization showed that WNT-calcium signaling is located in the nephrogenic zone; however, FACS analysis showed activity is located in the ureteric bud (UB) cells. This indicates WNT-Ca2+ signaling is active at the UB tips and could be involved in branching morphogenesis. Until now, WNT11 was thought to contribute to UB branching through paracrine activation of branching factors. Using kidney explants, we have demonstrated that WNT11 is likely acting in an autocrine fashion on UB tip cells to upregulate c-ret, a cell surface receptor required for normal UB branching. To determine WNT-Ca2+ pathway involvement in ADPKD, we investigated the levels of pathway activation in Pkd2/WS25 mice bearing the NFAT-Luc reporter gene. NFAT-Luc activity was drastically decreased in Pkd2/WS25 mouse kidneys as compared to wild-type littermates. We hypothesized that a further inhibition of the WNT-Ca2+ pathway during key points in embryonic kidney development could accelerate cystic development. This hypothesis was tested by oral gavage of pregnant female mice with the known WNT-Ca2+ pathway inhibitor cyclosporin-A (CSA). CSA treated Pkd2/WS25 mice showed a worse cystic phenotype at both post natal day (P)1 and P7 than control treated littermates. Studies are underway to ascertain whether expression of a constitutively active NFAT transgene corrects ADPKD. / En 2005, Simons et al. ont émis l'hypothèse d'un passage de la signalisation canonique à la signalisation non-canonique de WNT durant le développement du rein. L'échec de cette transition aurait pour conséquence le développement de kystes rénaux. Notre laboratoire a montré que la signalisation canonique de WNT est élevée pendant le développement du rein et est réprimée une fois la néphrogenèse accomplie dans les souris kystiques et de type sauvage. Nous supposons que la signalisation non-canonique de WNT pourrait être anormalement régulée dans les cas de polykystose rénale type dominant (ADPKD). Nous avons étudié le rôle de la voie non-canonique de WNT-calcium au cours du développement du rein ainsi que son implication potentielle dans l'ADPKD. Nous avons utilisé une souris possédant un gène rapporteur NFAT-luciférase, où NFAT est le facteur de transcription le mieux caractérisé en aval de la voie WNT-Ca2+. Ce gène rapporteur répond aux molécules non-canoniques de WNT, Wnt5a et Wnt11, et l'effet des WNTs pourrait être annulé par la cyclosporine A (CSA), un inhibiteur spécifique de cette voie de signalisation. L'hybridation in situ a prouvé que la voie de signalisation WNT-Ca2+ se trouve dans la zone de formation des néphrons. Toutefois, l'analyse par FACS a montré que l'activité de la luciférase se situe dans les cellules du bourgeon urétéral (UB). Ceci indique que la voie WNT-Ca2+ est activée aux extrémités de l'UB et pourrait être impliquée dans la morphogenèse par ramification de ce dernier. Jusqu'à présent, on pensait que Wnt11 participait à la ramification de l'UB via une activation paracrine. En utilisant des explants de rein, nous avons démontré que Wnt11 agi de manière autocrine sur les cellules à l'extrémité de l'UB pour activer c-ret, un récepteur membranaire nécessaire à la ramification normale de l'UB. Afin de déterminer l'implication de la voie WNT-Ca2+ dans l'ADPKD, nous avons étudié les niveaux d'activation de cette dernière dans les souris Pkd2/WS25, qui possèdent le gène rapporteur NFAT-luc. L'activité NFAT-luc est considérablement diminuée dans les reins de souris Pkd2/WS25 par rapport à celle des souris de type sauvage. Nous faisons l'hypothèse qu'une nouvelle inhibition de la voie WNT-Ca2+ pendant des moments cruciaux du développement embryonnaire rénal pourrait accélérer la formation de kystes. Cette hypothèse a été validée en gavant des souris gravides avec de la CSA. On a observé chez les souris Pkd2/WS25 traitées à la CSA un phénotype aggravé aux jours P1 (jour post-natal 1) et P7. Les contrôles ont été réalisés dans la même portée. Des études sont actuellement en cours pour déterminer si l'expression d'un transgène NFAT constitutivement actif corrige l'ADPKD.

Recent studies indicate that children conceived using assisted reproductive technologies are at increased risk of growth disorders and genomic imprinting diseases. Reports have suggested that these disorders could result from epimutations including abnormal DNA methylation patterns in imprinted genes. DNA methylation, catalyzed by a family of DNA methyltransferases (DNMTs), is both heritable and reversible and is the major epigenetic mark associated with imprinting. We propose that techniques used in ART, such as ovarian stimulation, and factors underlying infertility might interact to disrupt DNA methylation. In this study, we investigated whether superovulation in combination with DNMT1o deficiency could induce more severe perturbations of imprinted methylation in mouse blastocysts, than ovarian stimulation alone. Female mice heterozygous for lack of oocyte-derived DNMT1o and Dnmt1o +/+ controls were subjected to moderate (6.25 IU PMSG/hCG) or high (10 IU PMSG/hCG) hormone doses and then mated with B6 (CAST7) males. At 3.5 days, ovulation sites were counted and blastocysts were collected. Methylation patterns of the imprinted genes Snrpn, H19 and Kcnq1ot1 were assessed by bisulfite sequencing from pools of 5 to 8 blastocysts. The high hormonal dose induced higher ovulation rates but also the lowest yield of blastocysts. The moderate and high doses were selected for the DNA methylation analysis. DNMT1o deficiency did not significantly perturb Snrpn methylation patterns in blastocysts of females injected with the moderate dose of hormones. H19 methylation patterns were more susceptible to perturbation, showing a preliminary indication of a slight gain of methylation on the maternal allele after superovulation at the moderate and high doses. The high dosage regimen showed evidence of perturbed Kcnq1ot1 methylation patterns in one of the pools of blastocysts with the paternal allele showing an unexpected gain of methylation. Follow-up studies will be conducted by transferring blastocysts into pseudopregnant females and examining the effects at later times in development (i.e. mid-gestation). The aim will be to determine whether superovulation and DNMT1o-deficiency are associated with abnormalities in genomic imprinting in postimplantation development. / Des études récentes indiquent que les enfants conçus à l'aide de procréation médicalement assistée (PMA) sont à risque accru de troubles de la croissance et de maladies à empreinte génomique. Des rapports ont suggéré que ces troubles pourraient résulter d'épimutations anormales au niveau de la méthylation de l'ADN dans les gènes imprimés. La méthylation de l'ADN, catalysée par une famille d'ADN méthyltransférases (DNMTs), est la principale marque épigénétique associée à l'empreinte. Nous proposons que les techniques utilisées dans le cadre de la reproduction assistée, telles que la stimulation ovarienne, et l'infertilité pourraient interagir de manière à perturber la méthylation de l'ADN. Dans cette étude, nous avons examiné dans quelle mesure la superovulation jumelée avec un déficit au niveau de la DNMT1o pourrait induire davantage de perturbations graves de la méthylation dans les blastocystes de souris, que la stimulation ovarienne seule. Des souris femelles hétérozygotes pour le manque d'ovocyte dérivé DNMT1o et les contrôles ont été soumis à une stimulation de l'ovaire à dose modérée (6,25 UI de PMSG / HCG) ou élevée (10 UI de PMSG / hCG) d'hormones et puis accouplées avec des mâles B6 (CAST7). À 3,5 jours, les sites d'ovulation ont été comptés et blastocystes recueillis. Les profils de méthylation des gènes à empreinte génétique Snrpn, H19 ainsi que Kcnq1ot1 ont été évalués à l'aide du séquençage au bisulfite à partir d'un pool de 5 à 8 blastocystes. La dose élevée hormonale induit des taux d'ovulation élevés, mais aussi le plus faible rendement de blastocystes. Nous avons sélectionné les doses modérées et élevées pour l'analyse de la méthylation d'ADN. La carence en DNMT1o n'a pas perturbé de manière significative les profils de méthylation dans le gène Snrpn au sein de blastocystes de femelles injectées avec la dose modérée d'hormones. Les profils de méthylation de H19 ont été plus sensibles aux perturbations, révélant une hyperméthylation sur l'allèle maternel après superovulation à la dose modérée et élevée. Le dosage élevé induit une plus grande perturbation des profils de méthylation de Kcnq1ot1 dans l'un des pools: l'allèle paternel a arboré un gain surprenant de méthylation.D'autres études seront menées en effectuant un transfert de blastocystes chez des femelles pseudo-enceintes plus tard dans le développement (c'est-à-dire à mi- gestation), afin de déterminer si la combinaison de la superovulation et de la carence en DNMT1o est associée à des anomalies dans les gènes à empreinte génétique après l'implantation de l'embryon.

Hydatidiform mole (HM) is a form of gestational trophoblastic disease characterized by hydropic degeneration of chorionic villi of the placenta with or without embryo. My project is aimed at finding a new gene for recurrent HMs (RHMs). When I started this project there wasonly one gene, NLRP7, known to be causative of this condition, Mutations in a second gene, KHDC3L, in patients with RHMs were discovered by Parry et al. 2011. Genetic heterogeneity in the causation of RHMs is shown by our group and many others and is also evident from the absence of mutations in both NLRP7 and KHDC3L in some patients with RHMs. This suggests the possible role of other genes in the causation of RHMs. Exome sequencing was carried out on 18 patients with RHMs and no mutations in NLRP7 and KHDC3L. Analysis of the Exome sequencing data revealed NSVs in many genes. Of these, five candidate genes had two defective alleles in one patient, each. These variants were co-segregating with the phenotype and predicted to be damaging. Sequencing of these genes in more patients with recurrent HMs and reproductive wastage is underway in our laboratory and may help in finding new genes for molar pregnancies. After the identification of KHDC3L, I sequenced this gene in 97 patients with recurrent HMs and reproductive wastage who are negative for NLRP7. I found two frameshift deletion mutations, including one novel, in three patients with RHMs. I performed haplotype analysis using the SNPs surrounding one mutation and found that the haplotype is shared between patients of three ethnic groups, suggesting a strong founder effect for this mutation across populations. Co-localization experiments on lympoblastoid cell lines from normal subjects showed that that KHDC3L co-localizes with NLRP7 to the Golgi and microtubule organizingcenter. / La môle hydatidiforme (HM) est une forme de maladie trophoblastique gestationnelle caractérisée par une dégénérescence hydropique des villosités choriales du placenta avec ou sans embryon. Mon projet visait à identifier un nouveau gène responsable des môles hydatidiformes récurrentes (MHR). Quand j'ai commencé ce projet, un seul gène, NLRP7, était connu pour être responsable de cette condition. Des mutations dans un second gène, KHDC3L, ont été ensuite découvertes par Parry et al. 2011 chez des patientes atteintes de MHR. L'hétérogénéité génétique dans l'étiologie des MHR a été démontrée par plusieurs équipes, y compris la notre, et ressort également de l'absence de mutations dans les deux gènes NLRP7 et KHDC3L chez certaines patientes atteintes de MHR. Ceci suggère l'implication d'autres gènes qui causent les MHR. Le séquençage des exomes a été réalisé sur 18 patientes atteintes de MHR mais ne portant aucune des mutations dans NLRP7 ou KHDC3L. L'analyse des données de séquençage des exomes a révélé la présence de variants non-synonymes dans de nombreux gènes. Parmi ceux-ci, cinq gènes candidats présentant chacun deux allèles défectueux. Ces variants co-ségrégeaient avec le phénotype anormal et sont prédits dommageables. Le séquençage de ces gènes chez les patientes atteintes de MHR et d'autres formes d'echecs de reproduction est en cours dans notre laboratoire et pourrait aider à identifier de nouveaux gènes responsables des grossesses môlaires. Après l'identification du gène KHDC3L, j'ai séquencé ce dernier chez 97 patientes atteintes de MHR et d'autres formes d'echec de la reproduction mais ne présentant pas de mutations au niveau du gène NLRP7. Deux délétions dont une nouvelle affectant le cadre de lecture de ce gène ont été identifiées chez trois patientes atteintes de MHR. L'analyse des haplotypes par SNP autour de la mutation commune du KHDC3L a révélé le meme haplotype chez les trois patientes de differentes ethnies suggérant ainsi un effet fondateur important de cette mutation dans différentes populations. Des expériences de co-localisation sur des lignées cellulaires lympoblastoides chez des sujets normaux ont montré que KHDC3L co-localise avec NLRP7 au niveau de l'appareil de Golgi et le centre organisateur des microtubules ou centrosome.

<p> Many neurologic diseases cause discrete episodic impairment. Study of the genes and mechanisms underlying these diseases has informed our understanding of the nervous system. Here we describe a novel episodic neurologic disorder, which we term familial cortical myoclonus (FCM). FCM is characterized by adult onset, slowly progressive, multifocal, cortical myoclonus, inherited as an autosomal dominant trait. On the basis of clinical, electrophysiological, and genetic data, FCM is nosologically distinct. We utilized genome-wide single nucleotide polymorphism genotyping, microsatellite linkage, and massively parallel sequencing to identify a mutation in the <i>gene <u> n</u>ucle<u>ol</u>ar protein 3</i> (<i> NOL3</i>) that likely causes FCM. <i>NOL3</i> is thought to bind to pro-apoptotic proteins and thereby repress apoptosis, but our extensive experimentation did not replicate these claims. In vitro, the <i>NOL3 </i> mutation leads to post-translational modification of NOL3 protein. We could not pinpoint the identity of the modification, but did find that this process is regulated by phosphorylation at residue T114. Finally, a proteomic screen for novel binding partners identified two candidates that modulate neuronal/astroglial differentiation. We hypothesize that the <i>NOL3 </i> mutation abrogates these interactions to cause FCM. This hypothesis will be tested with <i>Nol3</i> mutant mice that we generated. In total, this work defines a novel episodic neurologic phenotype and the associated mutation, calls into question some of the published functions of <i> NOL3,</i> and presents an alternative mechanism that may explain the pathophysiology of FCM.</p>

The Drosophila melanogaster gene yuri gagarin was first identified in a screen for aberrant behavioral responses to the mechanosensory stimulus provided by the force of gravity. This novel gene encodes Yuri isoforms of four sizes; the three larger isoforms are predicted to be composed largely of coil-forming domains, which are common protein interaction domains. The four isoforms are expressed in varying ratios throughout the animal, at all stages of development. Notably, Yuri is present in muscle tissue, localized to Actin-containing regions of the sarcomere. A male-sterile deletion allele of yuri, yuriF64, was identified and sequenced. This deletion removes roughly 500 base pairs of sequence upstream of the Yuri coding region; in these animals, the mid-sized Yuri isoforms are not expressed in any tissue, and the smallest isoform is no longer expressed in the testis. In wild-type animals, Yuri protein is present throughout the testis, and it assumes a dynamic microtubule-dependent nuclear localization pattern during the process of nuclear condensation, appearing to be a component of the "dense complex" described in ultrastructural studies of spermatogenesis. A previously undescribed filamentous Actin network co-localizes with Yuri on the nuclear surface; Yuri may act as a linker between the Actin and microtubule cytoskeletons. In yuriF64 homozygotes, no Yuri-staining or filamentous Actin structures are detected on the condensing sperm nuclei, and the basal body and centriolar adjunct are mis-localized or absent. Axonemal defects are apparent in heterozygotes and homozygotes, and the Actin-dependent sperm individualization does not take place, leaving immature sperm bound in bundles of 64. Yuri physically interacts with components of the Actin cytoskeleton (Actin, Tropomyosin 1, and Troponin T) in an in vivo pull-down assay. Tropomyosin 1 (Tm1) is the interacting protein purified in greatest abundance in the assay, suggesting that it and Yuri interact directly; this coil-forming molecule wraps actin filaments, stabilizes them against enzymatic or physical degradation, and blocks the interaction of the filament with other proteins. Tm1 is present on the condensing sperm nucleus, and Yuri co-localizes with it in muscle tissue. Thus, the defects seen in yuriF64 animals may reflect aberrant Actin dynamics arising from a lack of Yuri-Tm1 interaction.

Body size in multicellular organisms is determined by the integartion of two factors: the rate of growth and the duration of growth. In most animals, the rate of growth is controlled cell autonomously by the insulin-stimulated Pi3 kinase (Pi3K) pathway. However, the duration of growth is controlled in a more complex manner that involves endocrine factors that act cell non-autonomously. For example, in insects such as Drosophila, the duration of each larval phase is regulated by the timing of release of the molting hormone ecdysone from the prothoracic gland (PG). The molecular mechanisms by which the rate of growth and the duration of growth are integrated remain poorly understood. To help shed light in this area, I have investigated the intracellular signaling events that regulate ecdysone release in the Drosophila PG. I have found that expressing activated Ras, or the targets of Ras signaling Raf or Pi3K, in the PG reduces fly size and accelerates larval development via precocious synthesis and release of ecdysone. In contrast, expression of dominant-negative (dn) Ras, Raf, or Pi3K increases fly size and prolongs larval development via delayed synthesis and release of ecdysone. These results indicate that Ras-Raf and Pi3K signaling act in the PG to regulate the duration of growth by altering the timing of ecdysone synthesis and release. Conversely, I have found that expressing dn-Ras or dn-Raf, but not dn-Pi3K, in cholinergic neurons increases fly size and prolongs larval development, whereas, expression of activated Ras or Raf, but not Pi3K, in cholinergic nerurons decreases fly size, but delays larval development. Inhibition of insulin signaling in flies, via chromosomal loss-of-function mutations, also decreases fly size and delays development, raising the possibility that Ras-Raf signaling in cholinergic neurons may affect fly size by controlling the rate of growth via systemic insulin signaling.

Microsatellites are repetitive DNA sequences with high rate of slippage mutations, which cause changes in length. As expected by the neutral theory of molecular evolution the level of polymorphism of these sequences is high when located in non-coding regions that may experience little selection. We tested the hypothesis that triplet repeat microsatellites located in coding regions of the social amoeba Dictyostelium discoideum should have a low level of polymorphism, given the presumably stronger effect on fitness of changes in coding DNA. We analyzed the length of 8 microsatellites located in coding regions of Dictyostellium discoideum in 114 clones of the North America population, divided in 5 subpopulations. Our results showed that each of the eight loci was very highly variable in the population. The lowest range of length variation was 17 repeats (51 bp) for a microsatellite in the ATG1 gene and the maximum range was 78 repeats (234 bp) for a microsatellite in the dimA gene. We tested the possibility that the level of polymorphism was due to population structure. Although present, the population structure was low, and consequently not responsible for the high polymorphism.

Plants must be capable of detecting and responding to a wide variety of environmental stimuli, yet much remains to be determined about the mechanisms by which these processes occur. I describe herein a proof-of-concept mutant screen for identifying mutants aberrant in their responses to a variety of stimuli including heat, cold, touch, and darkness. Such mutants will undoubtedly serve as tools to investigate the molecular pathways by which plants detect and respond to their environment. In addition, I also describe a large-scale gene expression experiment utilizing Affymetrix ATH1 microarrays which focuses on the short-term responses of plants to touch and darkness. These chips enable an expression level survey of over 22,000 genes simultaneously. We identify 589 genes that have increased expression in plants 30 minutes after touch stimulation and 171 genes that have decreased expression. We also identify 461 genes that have increased expression in plants after 30 minutes of darkness treatment and 72 genes that have decreased expression. The identities of the regulated genes suggest that calcium and kinase signaling, cell wall modification, disease resistance and secondary transcriptional responses may be altered in plants subjected to mechanostimulation or darkness.

Auxins are hormones important for numerous processes throughout plant development. Plants use several mechanisms to regulate levels of the auxin indole-3-acetic acid (IAA), including the formation and hydrolysis of amide-linked conjugates that act as storage or inactivation forms of the hormone. Certain members of an Arabidopsis amidohydrolase family hydrolyze these conjugates to free IAA in vitro. To examine the in vivo importance of auxin-conjugate hydrolysis, I generated a triple hydrolase mutant, ilr1 iar3 ill2 , which is deficient in three of these hydrolases. I compared root and hypocotyl growth of the single, double and triple hydrolase mutants on IAA-Ala, IAA-Leu, and IAA-Phe. The hydrolase mutant phenotypic profiles on different conjugates reveal the in vivo activities and relative importance of ILR1, IAR3, and ILL2 in IAA-conjugate hydrolysis. In addition to defective responses to exogenous conjugates, ilr1 iar3 ill2 roots are slightly less responsive to exogenous IAA. The triple mutant also has a shorter hypocotyl and fewer lateral roots than wild type on unsupplemented medium. As suggested by the mutant phenotypes, ilr1 iar3 ill2 imbibed seeds and seedlings have lower IAA levels than wild type and accumulate IAA-Ala and IAA-Leu, conjugates that are substrates of the absent hydrolases. These results indicate that amidohydrolases contribute free IAA to the auxin pool during germination in Arabidopsis. The regulation of IAA-conjugate hydrolase gene expression remains unknown. I have characterized the IAA-Leu-resistant mutant, ilr3-1, and cloned the defective gene, which encodes a basic helix-loop-helix leucine zipper transcription factor. ILR3 may modulate IAA-conjugate hydrolysis by regulating hydrolase expression or activity. Previous work on the IAA-Ala-resistant mutant iar1, which is defective in a predicted membrane protein similar to metal transporters, suggests a link between metal homeostasis and IAA-conjugate sensitivity. To gain insight into the function of IAR1, I conducted a genetic modifier screen to isolate second-site mutations that restore IAA-conjugate sensitivity to iar1. Here, I characterize one such mutant, mtpc2-1, and identify the gene defective as encoding a metal transport protein. The work described in this thesis suggests roles for multiple components in IAA-conjugate sensitivity in Arabidopsis, including conjugate hydrolysis, transcriptional regulation of relevant genes, and metal homeostasis.