The neuromuscular junction is the site where signals are transmitted from a nerve to a target muscle fiber. The mechanisms responsible for the maintenance of motor nerve terminals at synaptic sites are not understood. Here, I investigated the role of target-muscle fibers in the maintenance of frog motor nerve terminals. Cutaneous pectoris muscle fibers were selectively removed and prevented from regenerating while leaving the motor innervation intact. The role of target muscle fibers in nerve terminal structure and function was examined. First, the maintenance of presynaptic activity in the absence of target was assayed with the activity-dependent dye FM1-43. I found that target-deprived nerve terminals maintain their presynaptic function of synaptic vesicle recycling for up to 5 months of target deprivation. These results indicated that the molecular machinery required for vesicular release is maintained in a functional state for long periods of target deprivation. Second, I quantified the stability of target-deprived nerve terminals using in vivo repeated imaging. I found that most target-deprived nerve terminals were remarkably well maintained for several months after muscle fiber removal. These data indicate that the cues that confer stability to frog motor nerve terminals reside outside the muscle fibers such as in the synaptic basal lamina or terminal Schwann cells. Destabilization observed at some nerve terminals after extended target-deprivation, could result from the turning over of the stabilizing cues. Finally, the molecular organization of target-deprived nerve terminals was analyzed. I found that the levels of two synaptic vesicle proteins, SV-2 and synaptotagmin were reduced in target-deprived nerve terminals when compared to intact neuromuscular junctions. Analysis of cytoskeletal proteins revealed that F-actin was located at discrete bands along synaptic sites that do not colocalize with synaptic vesicle clusters. F-actin is suggested to be located at either the Schwann cell processes and/or the nerve terminal immediately above them. A possible adhesion between nerve terminals and Schwann cell processes, could contribute to the maintenance of the frog nerve terminal at the synaptic site. Finally, all target-deprived synaptic sites were found to be associated with variable levels of agrin immunoreactivity, implicating agrin as a possible maintenance molecule.

Animals exhibit physiological and behavioral changes in anticipation of and response to changing environmental conditions. One example of this is the influence of photoperiod on the reproductive cycle of temperate zone seasonal breeders, such as Syrian hamsters. In addition to physiological changes which prevent ovulation and reduce gonadal steroid secretion, photoperiod can influence behavioral responses to gonadal steroid replacement, suggesting that photoperiods which inhibit reproduction alter neural responsiveness to estradiol and progesterone. The experiments described here were designed to address possible mechanisms by which photoperiod might produce these changes. In Experiment 1, immunocytochemistry was used as a semi-quantitative and anatomically specific technique to explore the hypothesis that reduced neural responsiveness to gonadal steroids was due to changes in the number of estrogen and progestin receptors in nuclei mediating hormonal effects on reproductive behavior and physiology. Exposure to a short photoperiod was associated with decreased progestin receptor immunoreactivity (Experiment 1B) but not with changes in estrogen receptor immunoreactivity (Experiments 1A, 1D), suggesting that neural responsiveness to estradiol is reduced in short photoperiods but that this effect is not due to changes in estrogen receptor number. Changes in responsiveness to estradiol are not dependent on peripheral effects of photoperiod on estradiol metabolism, as progestin receptor-immunoreactivity was reduced in short days even when estradiol was implanted directly into the mediobasal hypothalamus (Experiment 1C). Experiment 2 utilized Fos as a marker of neuronal activity to identify specific neural sites where photoperiod might act to influence neural responsiveness to steroid hormones and sociosexual cues responsible for induction of lordosis. Effects of photoperiod on lordosis and Fos were highly variable among experiments, but exposure to a short photoperiod was associated in three of the four experiments (Experiment 2A, 2B, and 2D) with increased Fos immunoreactivity in the ventromedial hypothalamus. Exposure to a short photoperiod may increase the responsiveness of the ventromedial hypothalamus to one or more hormonal or sociosexual cues to which hamsters were exposed.

Several genes implicated in Parkinson's disease (PD) encode components of the ubiquitin-proteasome pathway. In a specific form of PD (human Autosomal Recessive Juvenile Parkinsonism, AR-JP), loss of functional Parkin (ubiquitin E3 ligase) results in a selective loss of midbrain dopaminergic neurons and a absence of Lewy bodies (LB) from the surviving dopaminergic neurons. Since cells in patients with AR-JP do not express functional Parkin, it is unclear why most neuronal and non-neuronal populations remain unaffected. One possible explanation is that most cells express a redundant ubiquitin E3 ligase(s) that is absent from dopaminergic neurons. Such candidate(s) redundant E3-ligase would be expected to fulfill several criteria: (1) bind similar E2 Ubiquitin conjugating enzymes; (2) interact with the same cellular substrates; (3) facilitate the formation of aggresome/lewy bodies with similar properties of those induced by Parkin; (4) be expressed in the nervous system but presumably absent (or largely absent) from dopaminergic neurons. In this thesis I have demonstrated that the Human Homolog of Drosophila Ariadne-1 (HHARI) is a candidate for such a redundant E3 ligase. In addition I have shown that even though HHARI induces the formation of LB like aggresomes in cell culture with properties similar to those produced by Parkin, these aggresomes differ in their detergent solubility properties. Using mouse C2C12 primary skeletal muscle cells with altered expressions of Ariadne-1 or Parkin, I determined if HHARI and Parkin may serve redundant protective roles. Using cell viability assays I have shown that HHARI does not confer protection to cells treated with toxic insults like those implicated in PD. On the contrary, using RNA silencing, I have shown that reduced Ariadne-1 expression appears to confer some benefit. Finally, based on phenotypes reported for Ariadne-1-/- escaper and Parkin-/- flies as well as our protein interaction data, I investigated the roles of Parkin and HHARI during myogenesis. Using engineered C2C12 cells I have shown that Ariadne-1 levels are tightly regulated in proliferating and differentiating C2C 12 cells and that increased cellular abundance of Ariadne-1 affects muscle terminal differentiation downstream of myogenin, strongly highlighting the importance of Ariadne-1 and perhaps the Ubiquitin Proteasome Pathway in myogenesis.

Synapses are the functional units of neuronal circuits and the sites of integration for multiple signaling pathways. Understanding the molecular basis of synaptic function is critical to understanding the bigger picture of how we think, learn, remember and how neurological diseases and disorders disrupt these faculties. Here we characterize the role of bendless (ben) in central synapse formation by utilizing the giant fiber system (GFS), a well established neuronal network in Drosophila melanogaster. Ben is an E2 conjugase and a key member of the enzyme cascade involved in ubiquitin dependent regulation. Ben was originally identified more than two decades ago and was believed to be involved in axon guidance as mutants lacked a synaptic connection between the giant fiber (GF) and its target, the jump motor neuron (TTMn). We have been able to redefine Ben function by demonstrating that an incipient GF-TTMn synaptic connection is present in ben mutants. We have also analyzed the synapse with the help of synaptic markers as well as studied its features at an ultrastructural level. By conducting cell autonomous rescue experiments we have spatially determined that ben has a presynaptic function in the GFS. We then used the TARGET system to temporally characterize the gene and have isolated a critical period for Ben function during development. We further assayed protein localization by generating GFP-tagged Ben constructs and have found the protein to be nuclear as well as cytoplasmic. Subsequent studies have identified two multifunctional proteins—Semaphorin1a and Distracted—to be putative targets of Ben action. We have also carried out a preliminary characterization of the synaptic roles other components of the ubiquitin system have in the GFS, such as the ubiquitin ligase highwire and the deubiquitinating proteases, fat facets and UBP2. In summary, we have found the ubiquitin conjugase Ben to have a novel and distinct role as a developmental switch in the establishment of a central synapse. The identification of likely downstream targets of Ben and comparison with related ubiquitin associated proteins suggest that a delicate regulatory balance has to be maintained in order for a synapse that is functionally and morphologically normal to be sculpted.

The zebrafish belladonna (bel) mutation was identified in a large-scale mutagenesis screen to identify genes involved in retino-tectal pathfinding in Tubingen, Germany. In bel mutants, after exiting the eye, retinal axons grow ipsilaterally instead of crossing the midline to form optic chiasm. bel mutants are semi-viable and live bel embryos at 5 days show a "dilated pupil" phenotype after which the mutation was named. Later work showed that bel mutants have functional eyes although the optokinetic response is reversed in the mutants. Previous work in our lab showed that most retinal axons in the mutants initially grow towards the midline but later turn ipsilaterally. Also, two major forebrain commissures, the anterior commissure (AC) and the post-optic commissure (POC) also failed to form in bel mutants. These studies showed that bel defects are restricted to forebrain. Detailed analysis of eye sections showed defects in bel eye morphology during embryonic and adult stages. Initial work also mapped the bel locus on chromosome 8 and finer mapping linked one z-marker on either side of bel locus (z24272 and z44909). My dissertation project was to clone the bel gene and understand its role in forebrain patterning and axon guidance. I identified that bel locus encodes a zebrafish lim-homeodomain transcription factor, Lhx2. To further understand how bel(lhx2) might affect axon guidance, I first showed that bel mutants have subtle defects in forebrain patterning in the regions where axons cross the midline. I also showed that these forebrain patterning defects lead to defects in expression of proper cellular and molecular axon guidance cues at the midline in bel mutants. Finally, I showed that bel(lhx2) is required for cell proliferation in the diencephalon. Thus my detailed analysis of bel mutants has revealed new roles for lhx2 in diencephalon patterning and axon guidance.

The ability of established synaptic connections to strengthen and weaken (synapse plasticity) underlies higher order behavior such as learning and memory. Several gaps remain in our understanding of the cellular and molecular changes in pre- and postsynaptic cells associated with synapse plasticity. Using the Drosophila larval neuromuscular junction as a convenient model system throughout my dissertation work, I have concentrated on understanding various molecular mechanisms that underlie this phenomenon. Specifically, I have addressed fundamental issues governing the postsynaptic cell in these processes, such as: (i) what are the crucial proteins required to scaffold the postsynaptic apparatus during plasticity? (ii) what are the molecules that are brought together at this scaffold and how are their levels regulated during plasticity? and finally (iii) what are the signals that initiate the formation of these scaffolds during plasticity? More specifically, I have uncovered a mechanism by which a novel synaptic protein, Guk-holder (Gukh), coordinates the formation of a tripartite complex at the synapse with two proteins containing multiple protein-protein interaction domains: Discs Large (DLG) and Scribble (Scrib) (Mathew et al., 2002). By bringing DLG and Scrib together at the synapse, Gukh likely coordinates the function of these two complexes. As stated above Fasciclin II (FasII) is one of several molecules that is localized to the synapse by Dig (Thomas et al., 1997a), and its levels and symmetric distribution at the synapse have been found to be important for regulating synapse plasticity (Ashley et al., 2005; Schuster et al., 1996b). I showed that FasII also undergoes fast recycling at the synaptic membrane and that this recycling is regulated by the Drosophila homolog of Amphiphysin (Amph)-(Mathew et al., 2003). Such a mechanism would be compatible with a potential activity dependent regulation of FasII levels during plasticity. To understand the initial signaling mechanisms by which synaptic scaffolds are organized during synapse formation, I further investigated the role of the Wingless receptor Dfrizzled2 (DFz2) in this process. Wg signaling has been previously implicated in synapse development and proper localization of postsynaptic proteins (Packard et al., 2002). In these studies I uncovered an unconventional mechanism involved in transducing the Wg signal in the muscle cell. I show that DFz2 is cleaved at the synaptic plasma membrane, and the c-terminus of the molecule traffics to the nucleus. This nuclear import of DFz2 is dependent upon Wg signaling and is important to transduce the downstream effects of Wg signaling at synapses (Mathew et al., 2005 in press). (Abstract shortened by UMI.)

As an excellent organism for genetic study and with the many useful genetic manipulation tools developed in the past two decades, Drosophila melanogaster is an important model organism that has been utilized to understand the genetic regulation of various developmental processes. Here, we focus on the development of the Drosophila CNS midline cells to pursue a long term goal to understand (1) the regulation of programmed cell death during nervous system development and (2) the interaction between nervous system and the mesoderm structures during early embryonic development. Programmed cell death is an essential part of the development of the nervous system for vertebrate and invertebrate animals (reviewed by Oppenheim 1991). In the Drosophila CNS midline, about 70% of the glia cells die during a critical period. We found that the elimination of unwanted midline cells during embryonic nervous system development is a very well coordinated process, which ensures that specific number of midline glia survive. The others are relocated and executed by proteolytic cascade, followed by engulfment and further degradation by macrophages. We observed that three cell death regulatory genes (rpr, hid and grim) in the 75C1,2 region of the 3rd chromosome are all expressed in midline cells that are chosen to be executed. Using several mutants and genetic deficiencies, we found that in order for the appropriate number of midline cells to be eliminated, the functions of multiple genes in the 75C1,2 region are required. To further understand this cell death process, we targeted the expression of rpr and/or hid to midline cells. We found that the two genes show synergism in inducing ectopic cell death, further suggesting that they may functionally interact to regulate the proper cell death pattern. Inside the embryo, the development of the nervous system is not a isolated process. The developing nerve cells interact with the mesoderm during early embryonic development. We found that the CNS midline cells are required for the differentiation of a specific group of mesoderm cells, the dorsal median cell. The well conserved epidermal growth factor (EGF) signaling pathway is at least partially responsible for this CNS midline regulation of mesoderm cell development. In addition, we found that the proper anterior/posterior location of the dorsal median cells depends on the function of wingless or patched segmental polarity gene, and that these two genes might act through regulating the spatial differentiation of the CNS midline cells.