Yun Zhang - US grants

DESCRIPTION (provided by applicant): We are interested to understand how individuals communicate with environment by olfaction and, particularly, how olfactory responses are shaped and modified by experience and environmental cues. To this end, we use a genetic model organism Caenorhabditis elegans and a new form of olfactory plasticity in this animal model to ask the following questions: What is the functional organization of the neuronal circuit underlying olfactory learning? How does experience generate modulatory cues to modify olfaction? How does the intrinsic property of the olfactory circuitry change in response to experience to generate olfactory learning? We previously showed that C. elegans learns to avoid the smell of pathogenic bacteria after ingestion of the pathogens. Serotonin signaling from a pair of serotonergic neurons ADF and an olfactory circuit are essential to direct this learning process. The physiological stress of infection enhances ADF serotonin signals through CaMKII and a Gq pathway; and the strengthened signaling promotes learning through a serotonin-gated channel in several interneurons. An olfactory neural circuit downstream of two pairs of sensory neurons is required for this learning to occur. We hypothesize that the enhanced serotonin signaling modulates the properties of the olfactory circuit, resulting in a change in olfaction. To test this hypothesis and characterize the molecular and cellular mechanisms of this learning process, we will first use genetic ablation and laser surgery to map the neural circuit underlying the olfactory learning and define the neuronal components of this network. Then we will use calcium imaging to examine the neuronal properties of the olfactory circuit in both na¿ve and learned animals. And finally we will use molecular genetics and biochemical tools to characterize how the training experience generates a neuromodulatory serotonin signaling through CaMKII and a Gq signaling pathway.

DESCRIPTION (provided by applicant): Learning is an essential function of the nervous system that allows animals to modulate behavior with adaptive values. While increasing amount of knowledge on the molecular underpinnings of learning provide insights into the mechanisms underlying learning, our understanding cannot explain behavioral changes in most learning paradigms. One major challenge of the field is to link the function of the underlying neuronal network with behavior and to address how the property of neural circuitry encodes learning. We use the genetic model organism C. elegans to address this question. In the past funding period, we have established a form of aversive olfactory learning whereby the nematode learns to avoid the smell of pathogenic bacteria that make it ill. This form of learning is analogous to Garcia effect, in which animals learn to avoid the smell or taste of a food that is associated with stomach distress. Using this learning paradigm, we have characterized the structure and function of the underlying neuronal network. Particularly, we show that a serotonergic neural circuit composed of the serotonergic neuron ADF and the downstream interneuron RIA, as well as motor neurons specifically regulate learned olfactory preference. The serotonin signal in ADF regulates the aversive learning on pathogenic bacteria. ADF responds to bacterial odors with increased intracellular calcium signals and the C. elegans homolog of CaMKII, UNC-43, acts in ADF to regulate learning. The postsynaptic neuron RIA is critically required for the aversive learning. RIA displays compartmentalized axonal activity that is correlated with head movement. Meanwhile, RIA axonal compartments also display synchronous activity that is evoked by olfactory stimuli. Interestingly, we show that the aversive training modulates the activity pattern of ADF and RIA in a way that is consistent with training-induced behavioral changes in olfactory preference. Thus, we hypothesize that these learning-correlated changes in the functional attributes of ADF and RIA neurons encode learning. We propose to test this hypothesis by characterizing the regulatory mechanisms and function of the learning correlates in ADF and RIA. We will first define the interaction between these two learning correlates by testing the possibility that the learning correlate in ADF regulates the learning correlate in RIA. We will als characterize the regulation of these learning correlates by examining the effect of several genetic factors that we have identified to mediate learning. We will also define the neurotransmission of ADF that regulates RIA activity. Second, we will characterize the function of the learning correlates in ADF and RIA. We will use molecular and optogenetics to manipulate the property of these neurons to (1) eliminate the training-induced changes in their activity patterns; and (2) build the learning correlates with genetic methods in the key neurons of the circuit, and then test the resulting effects on olfactory learning. These studies will revea how experience modulates the function of a neural network and leads to experience-dependent behavioral changes.

To survive, animals have to optimize their physiological and behavioral responses based on specific environmental cues. Through conserved signaling mechanisms, the insulin/insulin-like peptide (ILP) pathway plays essential roles in this process by regulating development, metabolism and life span in response to internal and external environments. Intriguingly, ILP signaling also regulates learning and memory, suggesting that ILPs act as a link between environment and neural function to generate optimal behavioral outputs. However, the molecular and cellular mechanisms through which ILPs regulate learning remain largely uncharacterized. Many animals, including humans, encode multiple ILPs in their genomes, such as the 40 ILPs in C. elegans, suggesting functional diversity and potential interaction among them. Recently, two C. elegans ILPs, INS-6 and INS-7, are shown to play opposite roles in regulating aversive olfactory learning. The observation that INS-6 inhibits ins-7 in this process and that the expression of these two ILPs are regulated by other ILPs suggest the existence of an ILP network that modulates behavior. Because INS-6 and INS-7 appear to coordinate the animal's behavioral responses with its physiological state, this further suggests that this network's regulation of learning involves environmental context. This R01 will test this hypothesis through the following: (1) define the ILP network composition and architecture that regulates learning; (2) map the cellular circuitry through which this network functions; and (3) demonstrate how environment modulates the activity of the ILP learning network. To address these aims, this R01 will involve high-throughput learning behavioral assays and in vivo calcium imaging of neuronal activities in different ILP mutant backgrounds and under different conditions. Finally, completion of this R01 will illustrate how the ILP network optimizes learning behavior to increase survival under different environments.

To survive, animals have to optimize their physiological and behavioral responses based on specific environmental cues. Through conserved signaling mechanisms, the insulin/insulin-like peptide (ILP) pathway plays essential roles in this process by regulating development, metabolism and life span in response to internal and external environments. Intriguingly, ILP signaling also regulates learning and memory, suggesting that ILPs act as a link between environment and neural function to generate optimal behavioral outputs. However, the molecular and cellular mechanisms through which ILPs regulate learning remain largely uncharacterized. Many animals, including humans, encode multiple ILPs in their genomes, such as the 40 ILPs in C. elegans, suggesting functional diversity and potential interaction among them. Recently, two C. elegans ILPs, INS-6 and INS-7, are shown to play opposite roles in regulating aversive olfactory learning. The observation that INS-6 inhibits ins-7 in this process and that the expression of these two ILPs are regulated by other ILPs suggest the existence of an ILP network that modulates behavior. Because INS-6 and INS-7 appear to coordinate the animal's behavioral responses with its physiological state, this further suggests that this network's regulation of learning involves environmental context. This R01 will test this hypothesis through the following: (1) define the ILP network composition and architecture that regulates learning; (2) map the cellular circuitry through which this network functions; and (3) demonstrate how environment modulates the activity of the ILP learning network. To address these aims, this R01 will involve high-throughput learning behavioral assays and in vivo calcium imaging of neuronal activities in different ILP mutant backgrounds and under different conditions. Finally, completion of this R01 will illustrate how the ILP network optimizes learning behavior to increase survival under different environments.

PROJECT SUMMARY/ABSTRACT micro RNAs (miRNAs) constitute a major regulatory mechanism of gene expression. Recently, whole-genome transcriptome analyses have implicated many brain-enriched miRNAs in various mental disorders. However, our understanding of how these disease-associated miRNAs regulate cellular and signaling events in the nervous system to mediate cognitive functions is very limited. The difficulty in addressing these questions is partly due to the high degree of the heterogeneity in the cell types of the brain regions involved in the diseases and the complex and partially understood mechanisms underlying the functional readouts employed in most studies. Here, we propose to use the genetic and genomic model organism C. elegans to investigate the neuronal functions of four highly important human mental disease-related miRNAs, mir-31, mir-128, mir-134, and mir-137, by characterizing the neuronal functions of the C. elegans homologues of these miRNAs. We found that all of the C. elegans homologues of these conserved miRNAs exhibit restricted expression patterns that overlap in an interneuron RIA. RIA regulates several fundamental neural functions, including learning. We propose to use RIA as a model neuron to study the function of these well conserved miRNAs in the nervous system. First, we propose to establish the paradigm for functional characterization of these mental disease-related miRNAs using one of the conserved miRNAs, C. elegans mir-269 that is the homologue of human mir-31, as a model. An initial analysis revealed that mir-269 regulates the intracellular calcium dynamics of RIA and a form of olfactory learning, in which RIA plays a critical role. Using whole-genome transcriptome analysis of isolated RIA neurons, we identified candidate genes for the direct targets of mir-269 regulation, which includes the C. elegans homologues of the mammalian mitochondrial uniporter and its regulatory protein. Based on these results, we propose that mir-269 regulates learning by regulating mitochondria-mediated intracellular calcium dynamics of RIA. We will combine molecular genetics, optical physiology and quantitative behavioral analysis to address this hypothesis in Aim 1. Second, after establishing the system and paradigm to characterize the neuronal function of miRNAs, we will analyze the function of all of the eight C. elegans homologues of mental disease-associated miRNAs that are expressed in RIA. We will leverage our expertise in characterizing neuronal activity and behavior to provide a set of mechanistic characterization for these clinically relevant miRNAs in a defined cell type using the functional readouts that have well-characterized molecular, cellular and circuit underpinnings. Given the high degree of conservation between C. elegans and human miRNAs, our findings using this tractable model will significantly advance our understanding of the fundamental role of these mental disease-associated miRNAs, which will help to uncover the pathology of many devastating neurological diseases and to guide future therapeutical studies on these human disease conditions.

PROJECT SUMMARY Alternative mRNA splicing (AS) is a fundamental process that regulates the expression of more than 90% of human protein-coding genes. The function of AS in the nervous system is particularly prevalent and has been implicated in multiple neurological disorders that impair learning. Although AS is implicated in activity- dependent gene expression underlying neural plasticity, no systematic analysis on AS has been performed on an in vivo learning model. The complexity of the mammalian brain poses challenges to this type of studies. Thus, we still do not understand (1) to what extent learning engages AS in the nervous system, (2) how AS contributes to learning-induced changes in neuronal gene expression, and (3) how learning modulates AS and splice isoforms of specific genes to generate learned behavior. Here, we propose to address these fundamental questions in C. elegans. The rationale is that the wiring and genetic make-up of the C. elegans nervous system are well characterized, dynamic gene expression can be profiled for the whole brain or individual neurons, functions of genes in learning can be dissected at the cellular resolution with genetic and imaging tools, and the fundamental properties of the development and function of the nervous system are well conserved between C. elegans and more complex animals. In addition, many forms of learning exhibited by C. elegans share similar behavioral characteristics and molecular underpinnings with those displayed by higher organisms. The overall goal of this project is to characterize how AS regulates learning and to provide insights into neurological defects in brain function under many disease conditions. The hypothesis of this project is that AS regulates neuronal gene expression to modulate neural function and produce learning. Specifically, we will first characterize the global patterns of AS network and splice isoforms that are regulated by a learning paradigm well-characterized in our laboratory. We plan to systematically analyze how learning alters splicing or isoform usage of all genes expressed in the C. elegans nervous system. Next, we will use genetic perturbations to address the causal function of learning-regulated splice isoforms of conserved molecules in neural activity and behavior. The grant is exploratory, because it (1) presents the first systematic analysis of AS in an in vivo model of learning and (2) introduces conceptual and technical advances to address causal links between AS and learning behavior. The proposed work is significant, because it (1) tests a highly plausible function of AS, a fundamental gene expression process conserved in eukaryotes, in learning, and (2) characterizes the mechanisms whereby AS of conserved molecules regulates neuronal gene expression and function to produce learning. Meanwhile, our grant is built on a substantial amount of preliminary results that support conceptual and technical productivity. The outcome of this study will provide critical and timely insights into the studies on AS in learning in other systems and advance understanding of learning defects in neurological diseases associated with aberrant AS.

PROJECT SUMMARY Learning is a complex process, and likely involves many areas of the brain that detect and process sensory inputs, integrate experience, and display behavior. Consistently, various neurological diseases that impair different brain areas are associated with profound defects in learning. Thus, bridging different spatial scales and understanding the dynamics of different brain regions are essential to understanding how learning occurs and potentially designing strategies to mitigate learning deficiency. However, it is currently not possible to achieve these goals in most experimental systems, and our understanding of learning is limited by the technical approaches by which either local circuit and cellular properties or coarse psychophysical parameters underlying learning are measured. Here, we propose to address these fundamental questions in a reduced system ? the nervous system of the nematode C. elegans. The rationale is that the wiring and genetic make-up of this network are well known, probing whole-brain dynamics with single-cell resolution with exquisite temporal resolution is technically ready for C. elegans, and the fundamental principles for the development and the function of the nervous system are well conserved between C. elegans and more complex animal models. Further, C. elegans exhibits many forms of learning, similar to those displayed by higher organisms in behavioral characteristics and molecular cellular underpinnings. Particularly, we will use an olfactory learning paradigm whereby C. elegans learns to avoid the odorants of pathogenic bacteria, a type of learning similar to the Garcia effect through which many animals, including humans, learn to avoid the smell and/or taste of a food that makes them ill. Our long-term goal is to understand how learning is encoded and executed by the function of the whole brain, and to inform the design of potential therapeutic strategies. The central hypothesis of this project is that learning engages global activity and the learned information is encoded in distinct functional modules. Specifically, we will test whether learned information is encoded in the learning-dependent changes in the activity patterns of individual functional modules and/or the interactions among the modules. To this end, we aim to image and analyze multi-cell and whole-brain dynamics under naive and learned conditions to characterize how learning alters the structure of the brain activities; further, we will introduce perturbations to the whole-brain dynamics and examine the consequences for learning. This work is innovative because (1) it brings a conceptual advance to understanding learning across scales, (2) it introduces technical advancement in whole-brain imaging and analyses, and (3) it demonstrates perturbation strategies for altering whole-brain dynamics that have behavioral consequences. It is significant, because it tests several highly plausible and likely conserved cellular and whole-brain dynamic models for learning and examine their behavioral consequences, it informs and facilitates learning studies in other systems, and it paves the way for designing interventions.