This research proposal addresses a key challenge to drug development: the paucity of biomarkers that reveal whether interventions affect implicated brain circuitry at early stages, in animals and humans, before embarking on lengthy and expensive clinical trials. Studies of humans and rodents have established sleep spindles, defining EEG oscillations of stage 2 non-rapid eye movement (NREM) sleep, as a mechanism of memory consolidation. A growing body of work implicates sleep spindle abnormalities in neurodevelopmental and neurodegenerative disorders characterized by memory impairment. In schizophrenia, sleep spindle deficits predict impaired sleep- dependent memory consolidation. Findings that increasing spindles via drugs or auditory or transcranial brain stimulation during sleep improves memory in healthy people, provides the impetus to target spindles to improve memory in disorders. But targeting spindles does not inevitably improve memory. Complementary rodent and human studies provide an explanation: sleep-dependent memory consolidation relies not on spindles alone, but on their precise temporal coordination with the other two cardinal NREM sleep oscillations: cortical slow oscillations (SOs) and hippocampal sharp-wave ripples. These findings make it clear that while spindles are promising targets for improving memory, (i) effective therapies need to increase spindles AND preserve or enhance their coupling with SOs and ripples, and (ii) to evaluate efficacy, we need new assays to identify spindles that couple with SOs and ripples to mediate memory versus those that do not. We propose to: (i) identify the most powerful translational measures of sleep spindles as assays of sleep-dependent memory consolidation (UG3), and (ii) to noninvasively manipulate them to compare their responses in healthy humans and rodents (UH3). Using invasive recordings in epilepsy patients and local field potentials (LFPs) in rats, we will first demonstrate that spindles that couple with both SOs and ripples (TriCS: triple-coupled spindles) are associated with memory consolidation, thereby validating TriCS as a translational biomarker of memory. We will then use machine learning to develop a classifier that identifies TriCS based solely on their scalp EEG features. We will validate the EEG spindle classifier by applying it to a dataset from healthy humans to demonstrate that TriCS, but not non-coupled spindles, correlate with memory consolidation. In both species, we will determine which spindle assay TriCS, SO-coupled spindles (SOCS) or total spindles predicts memory best. Finally, we will noninvasively manipulate the spindle assays in humans and rats. Genetic studies are implicating specific pathophysiologic mechanisms of spindle deficits in schizophrenia and autism and identifying novel targets and treatments. The rodent and human spindle assays that we will develop will facilitate the translation of these advances to the clinic by allowing the efficient evaluation of potential interventions early in the treatment development pipeline and the identification of the most promising candidates for clinical trials.
A fundamental challenge to finding effective treatments for neuropsychiatric disorders is the paucity of biomarkers that can gauge whether a potential intervention affects abnormal circuitry early in the treatment development pipeline, before embarking on expensive and lengthy clinical trials. This research proposal will develop novel physiological assays of sleep-dependent memory consolidation in humans and rodents that can be used to predict the efficacy of newly emerging treatments. These assays will enable the efficient screening and identification of treatments that can alleviate highly disabling cognitive deficits in schizophrenia.
