No abstract

Highlights d Synaptic inputs onto mPFC L2/3 pyramidal neurons are reduced in Cntnap2 KO mice d The frequency and amplitude of mEPSCs are reduced in the mPFC of Cntnap2 KO neurons d Decreased density of dendritic excitatory and inhibitory synapses in Cntnap2 KO mice d Phase-modulated spiking to slow LFP oscillations is altered in Cntnap2 KO units

No abstract

Loss of function mutations in CNTNAP2 cause a syndromic form of autism spectrum disorder (ASD) in humans and produce social deficits, repetitive behaviors, and seizures in mice. Yet, the functional effects of these mutations at the cellular and circuit level remain elusive. Using laser scanning photostimulation, whole-cell recordings, and electron microscopy, we found a dramatic decrease in functional excitatory and inhibitory synaptic inputs in L2/3 medial prefrontal cortex (mPFC) of Cntnap2 knock-out (KO) mice. In accordance with decreased synaptic input, KO mice displayed reduced spine and synapse densities, despite normal intrinsic excitability and dendritic complexity. To determine how this decrease in synaptic inputs alters coordination of neuronal firing patterns in vivo, we recorded mPFC local field potentials (LFP) and unit spiking in head-fixed mice during locomotion and rest. In KO mice, LFP power was not significantly altered at all tested frequencies, but inhibitory neurons showed delayed phase-firing and reduced phase-locking to delta and theta oscillations during locomotion. Excitatory neurons showed similar changes but only to delta oscillations. These findings suggest that profound ASD-related alterations in synaptic inputs can yield perturbed temporal coordination of cortical ensembles.All rights reserved. No reuse allowed without permission.

Hippocampal networks form maps of experience through spiking sequences that encode sensory cues, space or time. But whether distinct rules govern the emergence, stability and plasticity of externally driven and internallygenerated representations remains unclear. Using two-photon calcium imaging, we recorded CA1 pyramidal populations across multiple days, while mice learned and performed an olfactory, delayed, working-memory task. We observed anatomically intermixed spiking sequences, comprised of 'odor-cells' encoding olfactory cues, followed by 'time-cells' encoding odor-specific delay time-points. Odor-cells were reliably activated across trials and retained stable fields over days and different delays. In contrast, time-cells exhibited sparse, unreliable activation and labile fields that remapped over days and extended delays. Moreover, the number of odor-cells remained stable, whereas time-cells increased over days during learning of the task, but not during passive exposure. Therefore, multi-modal representations with distinct learning-related dynamics and stability can co-exist in CA1, likely driven by different neurophysiological and plasticity mechanisms.