Hyperexcitability in Alzheimer's Disease

ABSTRACT In the past funding period we focused on hyperexcitability, because several studies in rodents and humans suggested it played an important role in AD pathophysiology. We showed for the first time that there is a type of epileptiform activity that occurs extremely early and is common in diverse mouse models of AD. The

epileptiform activity was a very large but brief spike in the EEG. This was surprising because many research studies have focused on seizures. However, IIS in mouse models of AD were far more common and occurred at younger ages than seizures, suggesting IIS deserved more attention. Indeed, it is debated how common

seizures are in human AD, and many clinical studies have reported IIS. In this proposal we focus on IIS for these reasons. We ask several fundamental questions: where in the brain do IIS originate, and does that they change with age? If IIS are selectively inhibited is cognition improved? Given hyperexcitability leads to

increased extracellular Aβ, if we inhibit IIS is Aβ reduced? What are the mechanisms underlying IIS? In Aim 1 we will address the hypothesis that IIS begin in the dentate gyrus (DG) based on data showing that the DG recording of the IIS is earlier and larger than area CA1 and overlying cortex. We will also ask if

other brain areas generate IIS as mice age and pathology progressively worsens. In Aim 2 we will use closed- loop optogenetics to silence the principal cells of the DG, granule cells (GCs) and do so only during IIS to ask if this selective, transient GC silencing improves memory and pathology. We will first target all GCs and then the

subset of adult-born GCs, because the young GCs may make a unique contribution to IIS. Aim 3 addresses mechanisms of IIS. We hypothesize that acetylcholine (ACh) is key because IIS primarily occur in rapid-eye movement (REM) sleep, a time when cholinergic input to the forebrain rises. Also, medial septum (MS)

cholinergic neurons appear to be overly active in sleep in mice that simulate AD. In early AD, an enzyme critical for ACh synthesis is elevated, which we also found in an AD mouse model. Another clue is that, in epilepsy, IIS are caused by a paroxysmal depolarization shift (PDS), which is a large, sudden depolarization of

principal cells that lead to a burst of action potentials. Therefore, we hypothesize that IIS in AD reflect a PDS, and this is due to actions of MS cholinergic neurons in REM sleep. Pilot data support the hypothesis because the muscarinic agonist carbachol induces a PDS-like event in GCs. A concurrent depolarization is also

required, which appears to be due to a carbachol-dependent inward current and the glutamatergic input from DG mossy cells (MCs) which we found are hyperexcitable in transgenic mice. Both the inward current and MC excitation of GCs appear to be greater in a mouse model of AD than wild type mice. If we are correct, selective

optogenetic inhibition of cholinergic input or MC input would prevent the GC PDS and IIS, and pilot data suggest this is indeed the case. Together the experiments will address reasons for early hyperexcitability in AD, a topic that has attracted widespread attention in the AD research community. The proposal will also address mechanisms and therefore

potential future therapeutic strategies.