Return to search

Parvalbumin-Positive Interneurons' Orchestration of Episodic Memory in Health and Disease

Our lives unfold in space and time—we are able to be aware not only of the present instant but also to recollect the past and imagine the future, and our memories tend to be not instantaneous snapshots but rather bear a temporal, sequential dimension. This faculty of time travel allows us to adjust our current actions in light of what we have previously learned and with respect to what we aspire to become. It depends upon faithful records of our personal experiences, termed episodic memory. While over the last century we have learned a great deal about the molecular changes that support this kind of learning, the circuit-level mechanisms with which the brain implements the formation of episodic memory remain to be discovered. Failures of episodic memory can be catastrophic, and unfortunately, such dysfunction is commonplace in a number of human pathologies. In the neuropsychiatric syndrome of schizophrenia, the capacity to form and utilize episodic memory is compromised, a state of affairs that likely contributes to the difficulty people with schizophrenia have adjusting their actions to meet desired goals.

Attempts to understand the pathogenesis of schizophrenia’s memory deficits at the molecular level have yielded frustratingly few leads, making circuit-level inquiries a rational next step. Utilizing a genetic mouse model of schizophrenia susceptibility (Df(16)A+/- mice), we have taken a three-pronged approach to the analysis of the circuit mechanisms and missteps of episodic memory. We first developed a behavioral model of episodic memory, a variation on classical ‘trace’ fear conditioning, which involves the formation of an association between an innocuous stimulus (conditioned stimulus, CS) and a temporally separate aversive stimulus (unconditioned stimulus, US). Next, we turned to a region of the brain known to be required for trace fear conditioning and implicated in the pathogenesis of schizophrenia, dorsal CA1 of the hippocampus. Because network coordination and plasticity in dorsal hippocampal CA1 relies heavily on the activity of soma-targeting, parvalbumin-positive interneurons (PV+ INs), we hypothesized that they may be mediators of the associations built during trace fear conditioning. We therefore sought to characterize their activity during temporal association learning in both wild-type (WT) and Df(16)A+/- mice using two-photon calcium imaging. We simultaneously recorded local field potentials in the contralateral dorsal hippocampus to pair the discrete transformations captured through imaging with information about more global states of hippocampal activity. Finally, we manipulated the activity PV+ INs during various epochs of freely-moving trace fear conditioning to test hypotheses regarding their necessity for trace fear conditioning in healthy and schizophrenia-susceptible mice.

We found that Df(16)A+/- mice have severe deficits in trace fear conditioning when compared to mice that do not carry their defining mutation. Calcium imaging of PV+, peri-somatic boutons in dorsal CA1 over the course of trace fear conditioning revealed a marked increase in the number of detected boutons that initiate activity during the presentation of the CS and that sustain their activity across the time gap preceding delivery of the US. This shift in activity was notably absent in recordings from Df(16)A+/- mice. Consistent with the observations of others, analysis of local field potentials indicated that successful learning was associated with modulation of amplitude and theta-phase relation in mid- and fast-gamma frequency oscillations. This modulation was compromised in Df(16)A+/- mice. Finally, we found that inhibition of PV+ INs during encoding in Df(16)A+/- mice restores their response to the CS to near-WT levels of fear expression.

Our results support the thesis that temporary downregulation of PV+ IN activity during encoding is essential for the formation of complex, hippocampus-dependent associations including temporal association memory. We suggest that this transient disinhibition may serve to allow for the generation of new pyramidal cell ensembles to represent the associated stimuli. The heightened, sustained inhibition observed during post-learning trials in the calcium imaging experiments is consistent with a transition of the PV+ INs into a role of stabilizing the fledgling memory trace during consolidation. Our results also support the hypothesis that in our model of schizophrenia susceptibility, impairments in learning complex associations may be due to the inability of PV+ INs to modulate their activity appropriately over the changing phases of memory formation. We identify PV+ INs as a promising therapeutic target for schizophrenia as we were able to restore behavior of the susceptible mice during our assay of temporal association memory. Further studies combining pharmacogenetic or optogenetic manipulations with calcium imaging and LFP recording could shed light on the mechanisms of these shifts in network plasticity and may help to identify new approaches to the treatment of the debilitating cognitive deficits associated with schizophrenia.

Identiferoai:union.ndltd.org:columbia.edu/oai:academiccommons.columbia.edu:10.7916/d8-y0d9-p514
Date January 2020
CreatorsBalough, Elizabeth Maier
Source SetsColumbia University
LanguageEnglish
Detected LanguageEnglish
TypeTheses

Page generated in 0.0022 seconds