Kai Zinn - US grants

The basic architecture of a nervous system is largely determined by the pathway choices made by neuronal growth cones during embryonic development. Although the behavior of mammalian growth cones has been extensively studied in tissue culture, little is known about the molecular mechanisms by which specific pathway choices are determined in vivo. Knowledge about these mechanisms would help to explain how the mammalian nervous system develops before and after birth. This is relevant to the eventual understanding of genetic diseases that affect the structure of the brain and of sensory and motor systems. The development of the insect segmental ganglia is a good experimental system in which to isolate and study pathfinding events in vivo, because these ganglia are composed of a very small number of neurons, each of which makes a unique, genetically programmed set of pathway choices. Furthermore, the basic ganglion architecture is conserved between species suitable for cell biology studies, such as the grasshopper, and species with well-developed genetics, such as Drosophila. A variety of data suggest that individual axons or axon bundles in insect segmental ganglia are differentially labeled by surface recognition molecules that are used for growth cone guidance. The cell surface proteins known as fasciclins are good candidates for such recognition molecules. The gene encoding one of these proteins, fasciclin I, has been isolated in both grasshopper and Drosophila, and mutation in the Drosophila gene has been identified. Although this mutation alone does not cause a visible alteration of the embryonic nervous system, embryos bearing both the fasciclin I mutation and a mutation in the Drosophila homolog of the c-abl oncogene have a lethal phenotype in which the nervous system is disrupted. The nature of the description, however, is not understood. In the experiments described in this proposal, the role of fasciclin I in the development of the embryonic nervous system will be studied by analyzing in detail the phenotypes of embryos lacking both fasciclin I and abl. This will be done by examining the structure of the nervous system in double mutant embryos using light-level immunohistochemistry and electron microscopic reconstruction. Double mutant combinations will also be made with other mutations in genes encoding proteins expressed in the nervous system, and their phenotypes similarly analyzed. Potential cell-adhesion activities of fasciclin I will be studied by expressing the molecule on the surface of tissue culture cells and studying the aggregation behavior of the transformed cells. In second part of the proposal, a genetic and a biochemical method to identify other potential neuronal recognition molecules are described. In the genetic approach, new molecules that may be involved in the pathway of action of fasciclin I will be identified by screening for mutations in other genes that can be mutated to an abl -dependent lethal phenotype. The biochemical approach utilizes a monoclonal antibody against a carbohydrate moiety shared by many insect neuronal surface proteins. Two of these proteins are fasciclins; the others are presently unidentified and could include other neuronal recognition protein species will then be generated and used to isolate cDNA clones encoding them.

The basic architecture of a nervous system is largely determined by the pathway choices made by neuronal growth cones during embryonic development. Although the behavior of mammalian growth cones has been extensively studied in tissue culture, little is known about the molecular mechanisms by which specific pathway choices are determined in vivo. Knowledge about these mechanisms would help to explain how the mammalian nervous system develops before and after birth. This is relevant to the eventual understanding of genetic diseases that affect the structure of the brain and of sensory and motor systems. The development of the insect segmental ganglia is a good experimental system in which to isolate and study pathfinding events in vivo, because these ganglia are composed of a very small number of neurons, each of which makes a unique, genetically programmed set of pathway choices. Furthermore, the basic ganglion architecture is conserved between species suitable for cell biology studies, such as the grasshopper, and species with well-developed genetics, such as Drosophila. A variety of data suggest that individual axons or axon bundles in insect segmental ganglia are differentially labeled by surface recognition molecules that are used for growth cone guidance. The cell surface proteins known as fasciclins are good candidates for such recognition molecules. The gene encoding one of these proteins, fasciclin I, has been isolated in both grasshopper and Drosophila, and mutation in the Drosophila gene has been identified. Although this mutation alone does not cause a visible alteration of the embryonic nervous system, embryos bearing both the fasciclin I mutation and a mutation in the Drosophila homolog of the c-abl oncogene have a lethal phenotype in which the nervous system is disrupted. The nature of the description, however, is not understood. In the experiments described in this proposal, the role of fasciclin I in the development of the embryonic nervous system will be studied by analyzing in detail the phenotypes of embryos lacking both fasciclin I and abl. This will be done by examining the structure of the nervous system in double mutant embryos using light-level immunohistochemistry and electron microscopic reconstruction. Double mutant combinations will also be made with other mutations in genes encoding proteins expressed in the nervous system, and their phenotypes similarly analyzed. Potential cell-adhesion activities of fasciclin I will be studied by expressing the molecule on the surface of tissue culture cells and studying the aggregation behavior of the transformed cells. In second part of the proposal, a genetic and a biochemical method to identify other potential neuronal recognition molecules are described. In the genetic approach, new molecules that may be involved in the pathway of action of fasciclin I will be identified by screening for mutations in other genes that can be mutated to an abl -dependent lethal phenotype. The biochemical approach utilizes a monoclonal antibody against a carbohydrate moiety shared by many insect neuronal surface proteins. Two of these proteins are fasciclins; the others are presently unidentified and could include other neuronal recognition protein species will then be generated and used to isolate cDNA clones encoding them.

DESCRIPTION (provided by applicant): This proposal concerns the molecular and cellular mechanisms that determine synaptic connectivity in developing nervous systems. It focuses on a set of Drosophila neuronal cell surface/signal transduction proteins that we have been studying for many years: the transmembrane receptor tyrosine phosphatases (RPTPs). The six Drosophila RPTPs are central regulators of axon guidance and synaptogenesis. They have been extensively investigated using genetics, but we still known relatively little about the signaling pathways in which they function. This proposal is directed toward identification and characterization of components of these pathways. Specific aims 1 and 2 concern new methods for identification of RPTP ligands and coreceptors. We have already defined a Ptp10D binding protein, Sas, using one of these methods, and will characterize Sas's roles in regulation of Ptp10D function in vivo. We have also shown that a signal from glia to neurons is required for axonal expression of a Ptp99A binding protein. We will follow up on these discoveries, and also continue our search for new binding proteins for four different RPTPs. Specific aim 3 describes genetic screens for components of RPTP signaling pathways. We recently defined a remarkable and unique protein trafficking phenotype in embryos lacking both Type III RPTPs, Ptp10D and Ptp4E. Tracheal cells need to make new apical membrane in order to create the lumens of tracheal tubes. In the double mutant embryos, apical proteins accumulate in large intracellular vacuoles ("bubbles"). Basolateral proteins are localized normally. Our results show that EGFR, Rho family GTPases, and Rab GTPases are involved in generation of this phenotype. We suggest that it occurs because EGFR and Rho activites are upregulated when the RPTPs are absent. This shifts the balance between endocytosis and exocytosis, favoring accumulation of apical proteins in fused endocytic vesicles at the expense of the luminal surface. We will examine whether some of these pathway components are also used for regulation of axon guidance by these RPTPs. We will use the tracheal phenotype as the basis for F1 and F2 genetic screens to find new proteins in the RPTP signaling pathways, and will evaluate the functions of the genes we discover in both neurons and tracheae. Specific aim 4 concerns a new method for identification of in vivo substrates of RPTPs. We will express "substrate trap" mutants of RPTPs in neurons and tracheae, purify tyrosine-phosphorylated proteins that associate with the traps, and identify these proteins by mass spectrometry. We will then characterize the roles of these putative substrates in the RPTP pathways using genetics, and reconstruct their phosphorylation and dephosphorylation in cell culture. PUBLIC HEALTH RELEVANCE: This is a basic research project to discover mechanisms involved in creation of neuronal circuits during development. Although the work is conducted in Drosophila, most of the genes we are studying have human counterparts. We hope to reveal general principles that will facilitate an understanding of how human brain wiring is controlled before and after birth. Knowledge about wiring mechanisms may help researchers to understand diseases in which neuronal connectivity patterns are altered. These include schizophrenia and autism.

This program will provide predoctoral training of students preparing for research careers in Molecular, Cellular, and Systems Neuroscience. It involves 26 faculty members from the Biology, Physics, Chemistry, and Engineering Divisions. It is a continuation of a program previously supported by NIH. Some research areas of special emphasis are: 1) neural development (control of cell fate, axon guidance and synaptogenesis in a variety of systems); 2) signal transduction mechanisms in neurons (sensory processing in the visual and olfactory systems of vertebrates and invertebrates, and synaptic transmission and plasticity in hippocampal neurons); 3) behavior (simple and complex behaviors in vertebrates, arthropods, and nematodes); 4) computational neuroscience (studies of single neurons, system of neurons, and whole organisms). The major components of our training activities are: 1) each student's individual research program under one or more faculty sponsors; 2) an organized curriculum of graduate courses; 3) preparation for qualifying examinations; 4) teaching activities; 5) an extensive and wide-ranging seminar program. Support is requested for 16 predoctoral trainees, who will be admitted to graduate study for a Ph.D. in Biology or in Computation and Neural Systems. Criteria for admission into the program include a strong motivation for a career in research and high quantitative ability. Our expectation that trainees will continue into productive research careers is supported by the records of previous trainees. Caltech has a strong commitment (at both the institutional and the Divisional levels) to increasing the representation of minorities in science. In the Biology program, we have made special efforts to attract exceptionally talented students from under-represented minority groups, and we have been quite successful in this effort in recent years. A number of these students are primarily interested in neuroscience research. The training faculty are located within several building clustered near each other on the Caltech campus. Multi-user facilities include DNA sequencing, oligonucleotide synthesis, peptide synthesis, protein expression and purification, monoclonal antibody production, a transgenic and 'knockout' mouse facility, and the Biological Imaging facility.

We seek to understand the molecular and cellular mechanisms that determine synaptic connectivity in developing nervous systems. Our work primarily focuses on motor axon guidance and synaptogenesis in the neuromuscular system of the Drosophila embryo. This system is attractive because it contains only 34 motoneurons and 30 muscle fibers, and the pattern of neuromuscular connections is genetically determined and essentially invariant between embryos. Most neural structures in mammals are much more complicated, and mammalian neurons are usually not individually specified. Nevertheless, many of the molecules and mechanisms involved in process outgrowth, axon guidance, and synaptogenesis are conserved between flies and humans. Thus, the results of our studies are likely to be relevant to an understanding of human neural development. We have concentrated on a group of cell surface/signal transduction molecules, the axonal receptor-linked protein-tyrosine phosphatases (RPTPs), and have shown that they control specific motor axon guidance decisions. Some of these RPTPs have mammalian counterparts with very similar structures (e.g., fly DLAR and human LAR), and RPTPs in vertebrates are also expressed in neuronal processes. We now wish to advance our understanding of the mechanisms involved in axon guidance by characterizing genetic and biochemical interactions among the RPTPs. We will focus on two RPTPs (DPTP99A and DLAR) that oppose each other's signaling pathways in which these and other RPTPs function. This will be done by examining interactions with the transmembrane proteins gp150 and Appl, which bind to and are substrates for fly RPTPs, and by conducting a biochemical a biochemical screen for new substrates using the recently developed 'substrate trap' method. We will also attempt to identify cell-surface ligands for RPTPs using expression cloning techniques. Finally, we will perform a genetic screen for new cell recognition/signaling molecules involved in axon guidance and synaptogenesis. This 'modular misexpression' P element screen is designed to identify genes for which over-expression in all muscles or all neurons produces axon guidance phenotypes. Some of the proteins encoded by which genes might function in RPTP signaling most are likely to be involved in different pathways and processes. We are especially interested in cell-surface proteins that control innervation of individual muscle fibers, a process that is still poorly understood.

DESCRIPTION (provided by applicant): This proposal concerns spastin and blue cheese beached (bchs), two genes identified in screens we conducted for genes involved in motor axon guidance and synaptogenesis in Drosophila larvae. We selected these genes for further study because they encode members of highly conserved but poorly understood protein families that have not been previously implicated in neural development. The two genes are not related to each other, but both encode proteins likely to be involved in protein trafficking within neurons and can be studied using similar methods. Both genes have orthologs or relatives affected in human genetic diseases. (1) The first gene, spastin, is the ortholog of a human gene affected in autosomal dominant spastic paraplegia (ADSP). spastin encodes an AAA ATPase. These ATPases are involved in catalyzing assembly and disassembly of protein complexes involved in vesicle trafficking, protein degradation, microtubule dynamics, and other processes. Analysis of an AAA ATPase sequence does not allow definition of the cellular process(es) in which it participates, however, so the targets of Spastin function are still unknown. Neuronal overexpression of spastin causes convergence of central nervous system (CNS) axons onto the midline, spastin loss-of-function (LOF) null mutant larvae display altered synaptic morphologies at their neuromuscular junctions (NMJs). They also have reduced evoked junctional potential (EJP) amplitudes, indicating that their NMJ synapses are abnormal. Spastin-null animals that survive to adulthood are unable to fly, walk poorly, and have a shortened lifespan. To further analyze Spastin function, we will complete the morphological and electrophysiological analysis of larval NMJs. We will also perform electrophysiological tests to examine why the mutant adults cannot fly and examine the adult brain for structural defects and neurodegeneration. We will examine the mechanisms involved in human ADSP by introducing mutations that cause spasticity in humans into the fly gene and determining if these act as dominant negatives in Drosophila. To define other components of the pathways(s) in which Spastin acts, we will perform an enhancer/suppressor genetic screen using a spastin gain-of-function (GOF) eye phenotype. Candidate genes emerging from the eye screen will be tested for modification of the spastin GOF axonal phenotype and for interaction with spastin LOF mutations. (2) The second gene, bchs, encodes a protein closely related to the founding member of the BEACH domain protein family: the human protein whose loss causes Chediak-Higashi syndrome (CHS), a lethal genetic disease characterized by immunological and neurological defects. Cells from CHS patients contain abnormal giant lysosomes, bchs overexpression in neurons produces a unique phenotype in which bulges form at the junctions between motor axon trunks and side branches. bchs LOF mutations cause adult neurodegeneration phenotypes in the brain and eye, and bchs flies have short lifespans. In bchs larvae, some motor axon pathways are abnormally thickened, suggesting that individual axons are swollen or that additional axons have joined the pathways. Bchs contains a FYVE domain, which binds to phosphatidylinositol-3-phosphate (Ptdlns3P). It is a vesicular protein that occasionally colocalizes with a fluorescent marker (GFP-2XFYVE) for Ptdlns3P-containing endosomes; however, most Bchs vesicles are distinct from GFP-2XFYVE vesicles, suggesting that they represent different compartments. To study Bchs, we will analyze its subcellular localization and determine the vesicular compartment(s) in which it functions. We will examine bchs LOF and gain-of-function (GOF) phenotypes in the larval neuromuscular system using antibody staining, electron microscopy, and electrophysiology. We will also search for enhancers and suppressors of a bchs GOF eye phenotype. Candidate genes from the eye screen will be tested for modification of the bchs GOF neuromuscular phenotype and for interaction with bchs LOF mutations.

DESCRIPTION (provided by applicant): Obesity is a complex disorder caused by an imbalance between food intake and energy expenditure. These processes normally are precisely regulated, so that body weight can remain constant over a long time, in spite of variable food intake and activity. This homeostasis suggests the presence of strong regulatory mechanisms, and most biological mechanisms have an underlying genetic basis. The isolation of the Drosophila adipose mutation in the 1960's demonstrated that Drosophila can become obese, since that mutation causes a doubling in the overall fat content of adult flies. The strength of Drosophila as an experimental organism lies not only in its amenability to large scale, fast, and cheap genetic screening, but also in many years of genetic analysis, climaxing with the complete sequence of its genome, which makes it relatively easy to identify and clone genes with interesting mutant phenotypes. The list of biological problems for which Drosophila has been used successfully as a model is impressive, including the identification of genes that regulate the circadian rhythm, aspects of behavior such as learning and memory, and nervous system development. Many of these genes have provided clues to the discovery of their mammalian homologs. The use of Drosophila in obesity research opens the prospect of large-scale genetic screens to identify genes responsible for appetite control, body weight regulation, and fat storage, as well as analysis of the underlying biological mechanisms. We have isolated a series of obese mutants of Drosophila. The specific aims of the project will be to characterize their phenotypes, clone the genes, analyze the affected biochemical pathways, and isolate suppressor genes to reverse the genetic obesity defects.

DESCRIPTION (provided by applicant): Caloric restriction (CR), a decrease in nutritional level without causing malnutrition, is well known to increase lifespan across species, from yeast to rodents. Ad libitum feeding can therefore be viewed as having a toxic effect on lifespan. Indeed, overnutrition in humans is one of the leading risk factors for age-related diseases and mortality. We propose to characterize the effects of overnutrition in Drosophila, and seek ways to mitigate them by finding mutants that show either enhanced sensitivity or resistance to overnutrition. The identification of molecular pathways involved should provide tools in the understanding of this phenomenon, by leading us to the downstream genes that manifest the lifespan changes. Examination of the pathological and physiological changes resulting from overnutrition in normal and mutant flies will cast light on their involvement in aging and age-related diseases. Most investigations have been directed at the effects of undernutrition as a beneficial factor in extending lifespan. This proposal represents an alternate approach, focusing on overnutrition to exaggerate deleterious effects, thus providing a sensitized system in which to discover methods of ameliorating them.

DESCRIPTION (provided by applicant): Genetic screens in Drosophila identified many of the cell-surface and secreted (CSS) proteins that are intensively studied today as regulators of axon guidance in both vertebrate and invertebrate systems. This proposal describes a genetic screen for CSS proteins that function as synaptic target labels in the embryonic/larval neuromuscular system. This system is ideal for examination of target labeling mechanisms, because it contains only 36 motor neurons and 30 muscle targets and has an invariant innervation pattern. Each identified motor neuron innervates a specific muscle fiber. Although many genes that regulate axon guidance in this system have been identified, we know very little about how individual muscle fibers are recognized as targets by motor axons. To address this problem, we first defined CSS proteins that cause axonal mistargeting when they are overexpressed on all muscle fibers. We did this by constructing a database of all genes in Drosophila that encode CSS proteins likely to be involved in cell recognition events. We then searched through all the existing collections of UAS (GAL4 binding site)-containing ('EP-like') element lines to find insertions immediately upstream of these CSS genes that could be used to confer tissue-specific, high-level expression by crossing them to GAL4 "driver" lines. We obtained EP-like insertions that can drive 410 of the 979 genes in the database, or over 40% of the putative cell recognition repertoire. We crossed each line to a pan-muscle GAL4 driver and examined F1 progeny larvae by antibody staining and confocal microscopy. We found 30 genes whose expression on all muscles causes high-penetrance axonal mistargeting phenotypes but does not perturb muscle structure. Six of the genes are in a specific family encoding proteins with extracellular domains containing leucine-rich repeats (LRRs), which are protein interaction modules. This proposal describes experiments to assess the functions of four LRR proteins that are expressed in muscles and appear to function as synaptic target labels, and to determine if the LRR family encodes additional target labels. The first specific aim concerns the Tartan (Trn) and Capricious (Caps) proteins. Loss-of- function phenotypes for trn and caps suggest that they function in a partially redundant manner in the embryo. In larvae, selective expression of Trn or Caps on muscle 12 only produces alterations in targeting specificity. We will determine the loss-of-function (LOF) larval phenotypes generated by knockdown of both Trn and Caps in a single muscle or in all muscles. We will also attempt to develop a method for labeling single motor axons in larvae, so that we can observe how genetic perturbations affect targeting of individual identified axons. Specific aims 2 and 3 concern two "new genes", CG14351/haf and CG8561. We have used genetic and RNAi analysis to show that the proteins encoded by these genes are required for the normal innervation of ventrolateral muscles. We will make null mutations in these genes and conduct a genetic interaction screen to find components of the CG14351/Haf signaling pathway. We will also determine whether CG8561, the ortholog of a mammalian IGF-1 binding protein, is a component of the insulin/IGF-1 signaling pathway. The final specific aim describes experiments to examine the entire LRR family to determine if it encodes other muscle target labels. To do this, we will make UAS-cDNA constructs and obtain or make RNAi lines for 41 LRR genes and assess their phenotypes in larvae. For all genes producing phenotypes, we will then make a map of their expression patterns in muscle fibers during the period of axonal outgrowth. This information will allow us to begin to combine LRR protein perturbations, knocking down multiple genes on specific muscles, in order to examine whether muscle fibers are labeled for targeting by expression of specific ensembles of LRR proteins. PUBLIC HEALTH RELEVANCE: This is a basic research project to discover mechanisms involved in creation of neuronal circuits during development. Although the work is conducted in Drosophila, most of the genes we are studying have human counterparts. We hope to reveal general principles that will facilitate an understanding of how human brain wiring is controlled before and after birth. Knowledge about wiring mechanisms may help researchers to understand diseases in which neuronal connectivity patterns are altered. These include schizophrenia and autism.

DESCRIPTION (provided by applicant): It is now clear that protein synthesis is required for animals to establish long-term memories. Misregulation of synaptic tranmission and protein synthesis plays an important role in many diseases. Until recently, it was assumed that all of the proteins required for all neuronal function were made in the cell body. The discovery of polyribosomes at the base of neuronal synapses suggested the possibility that proteins might be synthesized in dendrites in response to synaptic activity. In the previous grant period, we focussed our attention on the development of imaging techniques to visualize protein synthesis in dendrites and discovered several forms of plasticity that are implemented by local protein synthesis. In this proposal we will examine the signaling mechanisms that couple miniature synaptic transmission (minis) to the protein translation machinery. We previously discovered that minis tonically inhibit the dendritic protein synthesis machinery. Loss of minis leads to an upregulation of translation and a rapid homeostatic response. We wish to examine which intracellular signaling pathways couple neurotransmitter receptor activation to the protein synthesis machinery. We will also determine whether there is stimulation-dependent assembly and trafficking of ribosomes in dendrites. Previous observations of ribosomes in dendrites of hippocampal neurons suggest that the translational capacity of synapses in spines is limited by the number of ribosomes available in the dendritic pool. Using biochemical approaches and dynamic time-lapse imaging, we will examine whether polyribosomes might be assembled locally in spines or trafficked to spines. One of big unanswered questions concerns the relative contributions of somatically vs. dendritic synthesized proteins to synaptic function and plasticity. Recently we have developed a technique that can be used to identify the constituents of the dendritically proteome. Building on this technology, we will modify our procedure for labelling proteins in lysates to include fluorescent labelling of proteins in intact cells and tissue slices. We will develop multiple fluorescent tags to separately track proteins made in the cell body and the dendrites.The fate of proteins synthesized in these two compartments will be analyzed over time to address the fractional contribution of somatic vs. dendritic protein synthesis to the synaptic protein population and how these contributions change with synaptic activity and plasticity.

DESCRIPTION (provided by applicant): The mammalian vascular system, lung, and kidney are branched tubular epithelial organs that transport gases or fluids. Although genes required for assembly of these organs have been identified, the developmental mechanisms that determine the shapes and sizes of their tubes are not well understood. The respiratory (tracheal) system of the Drosophila larva has provided a useful genetic model for the study of the development of complex branched tubular networks. Branching morphogenesis in the embryonic tracheal system is controlled by patterned interactions between a fibroblast growth factor (FGF) receptor tyrosine kinase (TK) ortholog, Breathless (Btl), and its FGF ligand, Branchless (Bnl). The developmental logic of the tracheal system is similar to that of the mammalian vascular system, where vascular sprouts expressing the vascular-endothelial growth factor (VEGF) receptor TK grow toward sources of VEGF. How are the airways in tracheal branches sculpted into the appropriate tubular shapes? We obtained an entry point into this problem when we discovered a unique tracheal phenotype caused by a double mutation eliminating both of the Type III receptor tyrosine phosphatases (RPTPs), Ptp4E and Ptp10D. The Ptp4E Ptp10D double mutation converts linear unicellular tubes into spherical cysts. Type III RPTPs are highly conserved regulators of receptor TK signaling, and we found that the phenotype involves the loss of negative regulation by the RPTPs of three growth factor receptor TK orthologs: epidermal growth factor receptor (Egfr), Btl, and Pvr (VEGFR ortholog). This phenotype may have never been found in earlier genetic screens because it is only observed when both Ptp4E and Ptp10D are mutated. There may also be no single component downstream of the RPTPs that could be mutated to generate such phenotypes, since the RTKs signal through many pathways. Thus, the identification of genes that regulate tube geometry may require a sensitized genetic screen based on the Ptp4E Ptp10D phenotype. This is the basis of the first specific aim, which describes a systematic screen for recessive mutations that confer enhancement or suppression of the phenotype. Because this is a time-consuming screen, requiring quantitative analysis of individual embryos using confocal microscopy, we will reduce the numbers of lines that need to be screened by using a 'phenotypic screening kit' of deletion (Df) mutations that we have defined. For each deletion that enhances or suppresses the phenotype, we will identify the responsible gene using insertion mutations and RNAi lines, which exist for most Drosophila genes. When we have mutations in individual genes in hand, we will examine their phenotypes in detail and analyze their epistatic relationships with each other, as well as with the RTKs and RPTPs, in order to define genetic pathways. The second specific aim describes a systematic approach by which we can localize and tag the protein products of genes identified in the screen. We can attach the proteins to fluorescent markers of various colors (for localization in live and antibody-stained preparations) and to epitope tags or enzymes (for biochemical characterization). This system will allow us to find proteins that are localized t the regions of cells where tube shape is controlled. We can also analyze tyrosine phosphorylation of the proteins and determine if they physically interact with each other in the embryo.

DESCRIPTION (provided by applicant): Leptin is a hormone that controls fat storage and energy expenditure. Leptin is made by adipocytes and acts on hypothalamic neurons that regulate energy balance. High leptin levels are an indicator of sufficient energy stores and should cause cessation of eating. However, most obese humans are leptin resistant. Leptin interacts with a JAK/STAT pathway-coupled receptor, and leptin resistance can be caused by downregulation of the pathway downstream of the receptor. If additional regulators of leptin receptor signaling in hypothalamic neurons could be identified, manipulation of their activities might provide ways to overcome leptin resistance. This proposal describes a new way to find such regulators using Drosophila genetics. Drosophila has dedicated adipocytes, and the activities of specific sets of brain neurons control fat storage and metabolism. The preliminary results reported here show that Drosophila fat content is regulated by a leptin-like JAK/STAT signaling pathway that acts in fat-regulating 'Fru' neurons. These findings show that flies and mammals use some similar genetic and neural mechanisms for control of fat storage. Drosophila is a model system that is amenable to fast and inexpensive forward genetic screening, and our results suggest that conserved JAK/STAT regulators identified in Drosophila would have relevance for research in mammalian systems. The objective of the first specific aim is to find proteins that affect fat content by regulating JAK/STAT signaling in Fru neurons. These will be identified by screening a set of ~200 genes identified as modulators of the JAK/STAT pathway in cultured cells, about 75% of which have human orthologs or relatives. Expression of each gene will be knocked down in Fru neurons using transgenic RNAi, and those genes for which knockdown affects triglyceride levels will be selected for further study. The identified candidate genes will be prioritized based on sequence relationships with human genes, effect size and direction, and availability of viable loss-of- function mutations and inserted elements that can be used for tagging. The objective of the second specific aim is to place the highest-priority subset of the regulators identified in Specific Aim 1 into their molecular and cellular contexts. Analysis of genetic epistasis will be used to determine whether the regulators act upstream or downstream of the receptor, and if they control the firing of Fru neurons. Regulators will be fluorescently tagged in vivo to reveal their cellular and subcellular expression patterns. The same tagging strategy will be used to create 'driver' lines that will confer gene expression in neurons that normally express the regulator, allowing manipulation of their activities. The expected outcome of the research proposed in specific aims 1 and 2 is the definition of a set of conserved modulators of JAK/STAT signaling that function downstream of the receptor in fat-regulating neurons. Negative regulators identified in these experiments might be potential drug targets whose inhibition could upregulate anorexigenic leptin signaling in leptin-resistant obese individuals.

? DESCRIPTION (provided by applicant): Understanding the computations that take place in brain circuits will require identifying the wiring diagrams of those circuits. In recent years seveal new methods have been developed to identify the brain's wiring diagrams. Each of these methods have some strengths and limitations. Importantly, there is no available anterograde monosynaptic tracer that can be used to regulate gene expression of synaptically connected neurons in species ranging from drosophila to mice. We propose to develop and validate a new genetically-encoded system to trace brain circuits by transsynaptic control of transcription that could overcome some of the limitations of the currently available strategies. We anticipate that this tool will open new opportunities for investigating the relationship between connectivity of neuronal circuits and brain function. The strategy that we propose is based on ligand- induced intramembrane proteolysis. In this system, neurons expressing an artificial ligand (emitter neurons) activate an engineered receptor on their synaptic partners (receiver neurons). Upon ligand-receptor interaction in synaptic sites, the engineered receptor is cleaved in its transmembrane domain and releases a protein fragment that regulates transcription in the synaptic partners. Our initial experiments in vivo, in transgenic drosophila, have confirmed the feasibility of this strategy as a method to record cell-cell interactions between neurons in the brain. We propose to optimize and validate this design towards identifying wiring diagrams of neuronal circuits in transgenic animals, both in mice and drosophila.

PROJECT SUMMARY/ABSTRACT Many cell surface proteins (CSPs) that are essential for neural development have been identified, but we still lack an overall understanding of the logic of the cell-cell interactions that program the assembly of neural circuits. Our long-term goal is to understand how cell-cell interactions mediated by CSPs program the assembly of the intricate synaptic patterns of nervous systems. Many years ago, it was proposed that in ?hard- wired? systems such as the fish optic tectum and the insect CNS and neuromuscular system, each neuron or neuronal type is labeled by ?identification tags? that control synaptic specificity, and that these tags are represented by specific CSPs called ?surface labels?. The original hypotheses predicted that surface labels that control synaptic specificity should be: 1) expressed on small subsets of neurons or muscles, 2) recognized by receptors whose expression is also restricted to small subsets of neurons (and might themselves be surface labels), 3) required for or influence the formation of specific synaptic connections, 4) encoded by families of related genes. We discovered a network of interacting CSPs that satisfies all of these criteria, using a new approach in which we selected proteins for in vivo analysis from a global in vitro interaction network. The Garcia group at Stanford and our group at Caltech generated an extracellular ?interactome? for all Drosophila immunoglobulin superfamily (IgSF) proteins, and found a subfamily of 21 2-Ig domain cell-surface proteins, the Dprs, that selectively binds to another subfamily of 9 3-Ig domain proteins, the DIPs. Each dpr and DIP gene is expressed by a small and unique subset of neurons, and mutations in these genes produce specific alterations in synaptic connectivity. The objectives of the present application are to define whether and how interactions between Dprs and DIPs constitute a ?connectivity code? that contributes to wiring specificity in the Drosophila larval neuromuscular system. The primary hypothesis underlying this application is that engagement of Dprs with their DIP partners provides information that can control synaptic targeting decisions. We plan to attain the objectives of this application through three specific aims. The first of these examines how Dpr-DIP interactions control formation of an axon branch of a specific motor neuron. The second analyzes how another DIP expressed on a single motor neuron controls innervation of its muscle target. The third creates tools for analysis of all Dprs and DIPs and identifies those expressed by specific motor neurons and muscles. The expected outcome of the proposed research will be the acquisition of new insights into the mechanisms by which interactions among CSPs control the specification of synaptic connections in a relatively simple model system. This will have a significant positive impact for human health by increasing our understanding of conserved mechanisms involved in nervous system development and disease in humans.

Project Summary/Abstract Many cell surface proteins (CSPs) essential for neural development have been identified, but we still lack an overall understanding of how cell-cell interactions mediated by these CSPs program assembly of complex neural circuits. Our long-term goal is to understand these processes. Many years ago, it was proposed that, in ?hard-wired? neural structures such as the fish retinotectal system and the insect optic lobe, each individual neuron is labeled by ?identification tags? that control synaptic specificity, and that these tags are represented by specific CSPs called ?surface labels?. The hypotheses predicted that surface labels that control synaptic specificity should be: 1) expressed on small subsets of neurons in each brain area, 2) recognized by receptors whose expression is also restricted to neuronal subsets, 3) required for formation of specific synaptic connections, 4) encoded by families of related genes. We discovered a network of interacting CSPs that satisfies all of these criteria, using an in vitro interaction screen of Drosophila CSPs. In this screen, we identified a subfamily of 21 2-Ig domain proteins, the Dprs, that selectively bind to another subfamily of 9 3-Ig domain proteins, the DIPs, forming a network called the Dpr-ome. In the visual system, neurons expressing a particular Dpr tend to be presynaptic to neurons expressing a DIP to which that Dpr binds in vitro. The objectives of the present application are to understand how Dpr-DIP interactions regulate competition among visual system neurons for neurotrophic signals, and to determine whether and how binding of a presynaptic Dpr to its postsynaptic DIP partner controls formation and function of synapses. The primary hypothesis underlying this application is that engagement of Dprs with their DIP partners provides information that influences cell fates and patterns of synaptic connections in the optic lobe. In particular, we hypothesize that trans-synaptic interactions between Dpr11 and its partner DIP-? are required for determination of cell numbers and specification of connections in the color vision circuit. Dpr11 is expressed by a subtype of UV photoreceptors, the yellow (y) R7s. The primary synaptic target for R7s is the amacrine neuron Dm8. DIP-? is expressed by a subset of Dm8s (?yDm8s?) that selectively arborizes with yR7s. yDm8s that do not successfully innervate R7s die. DIP-? controls their ability to compete for Dpr11-expressing yR7 targets and thereby regulates cell death. We plan to attain our objectives through two specific aims. Aim 1: Control of competitive interactions among Dm8s by Dpr11 and DIP-?. Aim 2: Control of synaptic selection by Dpr11-DIP-? interactions. The expected outcome of the proposed research will be the acquisition of new insights into the mechanisms by which interactions among CSPs control the assembly of neural circuits in the developing visual system. This will have a significant positive impact for human health by increasing our understanding of conserved mechanisms involved in development and disease.