It has long been thought memories leave a physical signature on the brain, but it remains unclear precisely what this signature might be. In the hippocampus, area CA1 is necessary for episodic memory, and as with other forms of memory the synaptic connections of CA1 neurons are attractive candidates for being key substrates of memory acquisition and retention. Most excitatory synapses arise on dendritic spines, which can appear and disappear within hours and undergo long-term potentiation or depression of their synaptic strength. Both learning and diseases that impair memory function can alter hippocampal spine density. To examine the idea spines serve as basic elements of memory storage, a goal of our work is to develop a time-lapse imaging technique capable of tracking individual hippocampal spines over weeks in live mice. We will then use this technique to watch the dynamics of spines in CA1 across the acquisition, retention, recall, and extinction of a hippocampal-dependent memory. Prior studies have compared hippocampal spines of naive versus trained animals by imaging fixed tissue. In vitro studies have also correlated neuronal activity to short-term spine dynamics in slice culture. However, these approaches did not permit studies of spine dynamics over many weeks, nor did they correlate spine dynamics to an animal's ongoing life experiences. To overcome these limitations, we will capitalize on a novel optical technology, two-photon microendoscopy that will provide us the first opportunity to track hippocampal spines longitudinally in live animals and to see how spine dynamics relate to episodic memory. This approach will allow us to address basic questions that have gone unanswered. Are hippocampal spines stabilized or destabilized during learning? Do spine dynamics during learning exhibit any observable spatial organization? For example, we will examine if new spines tend to arise clustered together along the same dendrites. Thus, our aims are:
Aim 1 : Develop a chronic mouse preparation for time-lapse in vivo microendoscopy imaging of CA1 hippocampal spine dynamics over weeks and months.
Aim 2 : Examine CA1 spine dynamics over the course of hippocampal-dependent contextual fear conditioning in live mice. We will track individual spines and their dynamics across baseline conditions, following fear conditioning, and after memory extinction. Since we will observe the same neurons before and after learning, we will be able to test for relationships between cellular changes and subjects'behavior as a function of time. Due to the differences between hippocampal and neocortical forms of memory, we expect CA1 spine dynamics will display temporal and spatial characteristics distinct from those of neocortical spines, which have been studied previously by intravital microscopy. In the future, our imaging methodology will be applicable to other deep lying brain areas and to the study of other sub-cellular elements, such as organelles or synaptic supramolecular structures that have also been implicated in learning and diseases of memory and cognition.
In neuroscience, a current limitation is the inability to visualize neurons that lie deep within the brains of living mammals. Our research seeks to create a brain-imaging technique that will allow researchers to observe such deep lying neurons in live mice and to track their sub-cellular properties over weeks and months in an animal's life, in both normal animals and animal models of brain disease. We will then use this imaging technique to examine whether dendritic spines in the hippocampus show dynamical properties consistent with their being an important cellular substrate for long-term memory storage and a key locus for diseases afflicting memory.