Harel Z. Shouval - US grants

Synaptic plasticity is believed to be an important mechanism contributing to of learning, memory, and many aspects of development. There is significant evidence that in cortex synaptic plasticity contributes significantly to receptive field development. For example, in the hippocampus there is abundant cellular and molecular information about long term potentiation (LTP) and long term depression (LTD), the cellular manifestation of long lasting synaptic plasticity. LTP and LTD can be induced by different induction paradigms that depend on presynaptic rate, on pairing presynaptic spikes with postsynaptic depolarization, and on the precise time difference between pre and postsynaptic spikes. We have recently hypothesized that a single model, which depends on calcium influx through NMDA receptors can account for these different induction paradigms. Here we propose a more detailed study of the molecular dynamics, including improved but simple models of CaMKII and Calcinurin that underlie synaptic plasticity. Based on this detailed study as well as new experimental results and measured parameters, we will develop an updated version of the unified plasticity model (UPM) that can be quantitatively tested. We hypothesize that fluctuations in molecular dynamics can play a significant role in the resulting synaptic plasticity. We propose to analyze these fluctuations and calculate their effect on the different induction paradigms of synaptic plasticity. We also propose to test experimentally the validity of a key assumption of the UPM, that the back propagating action potential has a long tail, and to measure key physiological parameters in hippocampal cells. We will use measured physiological parameters in hippocampal cells in simulating the UPM in order to create a quantitative theory appropriate for the hippocampus. The UPM will be further developed to account for the maintenance phase of synaptic plasticity. We will also test the hypothesis that homeostatic metaplasticity is crucial in attaining stable, selective and robust fixed points.

DESCRIPTION (provided by applicant): We all have memories that date back to our youth; we remember the house we lived in at age 4; we remember a favorite schoolteacher. The mechanism for storing these memories is believed to be the long-term plasticity of synaptic connections within specific neuronal circuits. However, this putative cellular basis of memory relies on proteins that typically have lifetimes far shorter than the memory. Here exactly lies a fundamental problem of long-term memory and synaptic plasticity: How can memories be stored for a human lifetime on the basis of proteins that are continuously degrading? Recently, it was shown that the brain-specific PKC isoform, protein kinase M??(PKM?), plays a unique role in maintaining both late long-term potentiation (L-LTP) of synapses and long-term memory. This crucial observation, however, does not explain how PKM??can overcome the natural degrading effect of protein turnover and diffusion. The central hypothesis of this proposal is that PKM??, through its control of its own synthesis, can form a bi-stable system, which can account for the maintenance of synapse specific long-term plasticity and memory. Here we propose to mathematically formulate this hypothesis within a biophysical model, and to analyze this model so as to propose testable experimental predictions. We then will directly test these predictions on PKM?-mediated persistent synaptic potentiation, using novel techniques tailored for testing the theory. Intellectual Merit: The finding that PKM??is both necessary and sufficient for the maintenance of synaptic plasticity and long-term memory has fundamentally changed the field of learning and memory, but much needs to be learned about the mechanisms that can actually accomplish the persistence of long-term plasticity and memory. This proposal addresses these questions using a combined theoretical and experimental approach. Such a theory in which bi-stability depends on regulation of translation is novel not only for neuroscience but also for biology in general. Our collaboration is uniquely qualified to carry out the proposed work because the Shouval lab has ample experience in modeling synaptic plasticity in collaboration with experimental groups, and the Sacktor lab has pioneered the science of PKM??and has ample experience with the proposed techniques. The experimental techniques include two new methodologies necessary for testing the predictions. First, we propose to test the model's predictions on protein translation in L-LTP, not by general protein synthesis inhibitors that may have issues of toxicity and indirect effects, but by use of antisense oligodeoxynucleotides directed to the translation start site of PKM??mRNA to specifically block PKM??synthesis in induction and maintenance. Second, because PKM?-mediated potentiation is both highly stable and yet rapidly reversible, we will use a fast-flow hippocampal slice chamber optimized for the study of the maintenance of L-LTP to test key predictions of the model. The proposed stochastic simulations of translation-dependent bi-stability are also novel in computational biology. Broader Impact: As the first demonstrated molecular mechanism of experience-dependent, long-term information storage in the brain, PKM??has significant clinical implications, and within the last year has been shown to contribute to in the biology of a variety of neurological and psychiatric diseases, including post-traumatic stress disorder, central neuropathic pain, and drug abuse. In order to assist the rapidly growing interest in PKM??in many labs, we will make our model accessible to the larger community, allowing for other scientists to test, modify, and incorporate their findings into the model, thus accelerating the pace of scientific discovery. Because an important goal for NSF is to integrate research and education, we will train a diverse pool of students. Our labs already train undergraduates, the Shouval lab takes undergraduates each summer through an REU program (PI S. Cox, Rice), and a UT system grant (PI H. Shouval), and local undergraduates throughout the year, and the Sacktor lab has had a long history of mentoring local disadvantaged high school students (e.g., through the Intel program). Both labs are dedicated to public outreach; for example, an article on PKM??and memory was on the front page of The New York Times. We are eager to extend this type of outreach to the domain of the interaction between theory and experiment in biological sciences.

Project Summary Abstract Learning new tasks and exposure to new environments lead to changes in the dynamics of brain circuits, as observed in various recent experiments. The ability to embed the statistics of the environment within brain circuits is essential for animals ability to thrive and survive in changing environments. However, the mechanisms by which circuits dynamics are implemented and learned are not well understood, and pose significant theoretical challenges. Recent work in both theoretical and experimental labs has highlighted the importance of circuit dynamics. Yet in most theoretical models the network connectivity is either not plastic, or obeys biologically implausible learning rules. Here we will develop a theory of how brain circuits can learn their dynamics from the statistics of the environment. We will anchor this work in a set of experiments, in order to make it biologically realistic and limited in scope. In aim 1 we will try to understand how networks can learn stimulus-reward spatiotemporal statistics. This aim will be based on circuit level experiments that show how neuronal dynamics change due to a stimulus followed by a delayed reward, and by cellular experiments that shed light on the mechanisms of reinforcement learning. This is a problem we know more about, and it is also inherently simpler than learning the statistics of the environment in an unsupervised manner. In aim 2 we will concentrate on experiments on which cortical circuits learn the order, but not the timing, of a spatiotemporal sequence. In such networks the timing of the learned sequence are determined by intrinsic network dynamics; making this problem simpler than learning both the order and the timing of a sequence. In aim 3 we develop networks and learning rules that can learn both the order and the timing of a spatiotemporal sequence. This effort will build on results in aim 2 in which the order of events is learned in an unsupervised manner, and of aim 1 in which the timing of events is learned using reinforcement learning.