Selective vulnerability of the ventral hippocampus-prelimbic cortex axis parvalbumin interneuron network underlies learning deficits of fragile X mice.

High-penetrance mutations affecting mental health can involve genes ubiquitously expressed in the brain. Whether the specific patterns of dysfunctions result from ubiquitous circuit deficits or might reflect selective vulnerabilities of targetable subnetworks has remained unclear. Here, we determine how loss of ubiquitously expressed fragile X mental retardation protein (FMRP), the cause of fragile X syndrome, affects brain networks in Fmr1y/- mice. We find that in wild-type mice, area-specific knockout of FMRP in the adult mimics behavioral consequences of area-specific silencing. By contrast, the functional axis linking the ventral hippocampus (vH) to the prelimbic cortex (PreL) is selectively affected in constitutive Fmr1y/- mice. A chronic alteration in late-born parvalbumin interneuron networks across the vH-PreL axis rescued by VIP signaling specifically accounts for deficits in vH-PreL theta-band network coherence, ensemble assembly, and learning functions of Fmr1y/- mice. Therefore, vH-PreL axis function exhibits a selective vulnerability to loss of FMRP in the vH or PreL, leading to learning and memory dysfunctions in fragile X mice.

Circuit mechanisms of navigation strategy learning in mice.

Navigation tasks involve the gradual selection and deployment of increasingly effective searching procedures to reach targets. The brain mechanisms underlying such complex behavior are poorly understood, but their elucidation might provide insights into the systems linking exploration and decision making in complex learning. Here, we developed a trial-by-trial goal-related search strategy analysis as mice learned to navigate identical water mazes encompassing distinct goal-related rules and monitored the strategy deployment process throughout learning. We found that navigation learning involved the following three distinct phases: an early phase during which maze-specific search strategies are deployed in a minority of trials, a second phase of preferential increasing deployment of one search strategy, and a final phase of increasing commitment to this strategy only. The three maze learning phases were affected differently by inhibition of retrosplenial cortex (RSC), dorsomedial striatum (DMS), or dorsolateral striatum (DLS). Through brain region-specific inactivation experiments and gain-of-function experiments involving activation of learning-related cFos+ ensembles, we unraveled how goal-related strategy selection relates to deployment throughout these sequential processes. We found that RSC is critically important for search strategy selection, DMS mediates strategy deployment, and DLS ensures searching consistency throughout maze learning. Notably, activation of specific learning-related ensembles was sufficient to direct strategy selection (RSC) or strategy deployment (DMS) in a different maze. Our results establish a goal-related search strategy deployment approach to dissect unsupervised navigation learning processes and suggest that effective searching in navigation involves evidence-based goal-related strategy direction by RSC, reinforcement-modulated strategy deployment through DMS, and online guidance through DLS.

Strategy updating mediated by specific retrosplenial-parafascicular-basal ganglia networks.

Adaptive behavior requires flexible control over learning and exploitation of potentially viable options. Within a particular task, careful learning of strategies that differ from the initially learned rule is especially important as it sets an individual's strategy repertoire. However, whether and how such strategy updating is mediated by specific brain networks has remained unclear. Retrosplenial cortex (RSC), a cortical area exhibiting extensive connectivity to dorso-medial striatum (DMS) and the hippocampal formation, has been broadly implicated in flexible learning and might be involved in strategy updating. Here, we investigate the specific role of mouse RSC in flexible learning, map relevant RSC-anchored cortico-thalamo-basal ganglia circuits, and dissect their role in strategy updating. Activity in RSC was neither required for initial rule learning nor to switch between previously learned rules but was specifically required to explore and learn new alternative options when previous ones were available but no longer appropriate. Such strategy updating depended on activity in RSC c-Fos+ ensembles associated with the original rule and on their connections to DMS and thalamic parafascicular nucleus (PF) neurons. At the circuit level, rule-related RSC projection neurons branched to innervate both DMS and PF neurons and mediated strategy updating through a RSC-DMS-substantia nigra reticulata (SNr)-PF network, coupling alternative exploration to outcome. In addition, a separate RSC-PF-RSC looped network promoted alternative exploration. Our results uncover cortico-basal ganglia-thalamo and cortico-thalamo networks involving subpopulations of neurons in RSC and PF that specifically control and implement strategy updating.

Absence of familiarity triggers hallmarks of autism in mouse model through aberrant tail-of-striatum and prelimbic cortex signaling.

Autism spectrum disorder (ASD) involves genetic and environmental components. The underlying circuit mechanisms are unclear, but behaviorally, aversion toward unfamiliarity, a hallmark of autism, might be involved. Here, we show that in Shank3ΔC/ΔC ASD model mice, exposure to novel environments lacking familiar features produces long-lasting failure to engage and repetitive behaviors upon re-exposure. Inclusion of familiar features at first context exposure prevented enhanced dopamine transients in tail of striatum (TS) and restored context-specific control of engagement to wild-type levels in Shank3ΔC/ΔC mice. Engagement upon context re-exposure depended on the activity in prelimbic cortex (PreL)-to-TS projection neurons in wild-type mice and was restored in Shank3ΔC/ΔC mice by the chemogenetic activation of PreL→TS projection neurons. Environmental enrichment prevented ASD-like phenotypes by obviating the dependence on PreL→TS activity. Therefore, novel context experience has a key role in triggering ASD-like phenotypes in genetically predisposed mice, and behavioral therapies involving familiarity and enrichment might prevent the emergence of ASD phenotypes.

[Formula: see text] Long-term verbal memory deficit and associated hippocampal alterations in 22q11.2 deletion syndrome.

Chromosome 22q11.2 deletion syndrome (22q11.2DS) is a genetic disease associated with an increased risk for schizophrenia and a specific cognitive profile. In this paper, we challenge the current view of spared verbal memory in 22q11.2DS by investigating verbal memory consolidation processes over an extended time span to further qualify the neuropsychological profile. Our hypotheses are based on brain anomalies of the medial temporal lobes consistently reported in this syndrome.Eighty-four participants (45 with 22q11.2DS), aged 8-24 years old, completed a verbal episodic memory task to investigate long-term memory on four different time delays. We compared trajectories of forgetting between groups (22q11.2DS vs. controls) and analyzed performance inside the 22q11.2DS sample through cluster analyses. Potential links between memory performance and volume of the hippocampal subfields were examined.We showed accelerated long-term forgetting (ALF) in the 22q11.2DS group, visible after a delay of one day. Using mixed models, we showed significant differences in the shape of memory trajectories between subgroups of participants with 22q11.2DS. These sub-groups differed in terms of memory recognition, intellectual functioning, positive psychotic symptoms and grey matter volume of hippocampal subfields but not in terms of age.In conclusion, by investigating memory processes on longer delays than standardized memory tasks, we identified deficits in long-term memory consolidation leading to ALF in 22q11.2DS. Nevertheless, we showed that a subgroup of patients had larger memory consolidation deficit associated with lower intellectual functioning, higher rates of positive psychotic symptoms and hippocampal alterations.

Long-Lasting Rescue of Network and Cognitive Dysfunction in a Genetic Schizophrenia Model.

Although sensitizing processes occur earlier, schizophrenia is diagnosed in young adulthood, which suggests that it might involve a pathological transition during late brain development in predisposed individuals. Parvalbumin (PV) interneuron alterations have been noticed, but their role in the disease is unclear. Here we demonstrate that adult LgDel+/- mice, a genetic model of schizophrenia, exhibit PV neuron hypo-recruitment and associated chronic PV neuron plasticity together with network and cognitive deficits. All these deficits can be permanently rescued by chemogenetic activation of PV neurons or D2R antagonist treatments, specifically in the ventral hippocampus (vH) or medial-prefrontal cortex during a late-adolescence-sensitive time window. PV neuron alterations were initially restricted to the hippocampal CA1/subiculum, where they became responsive to treatment in late adolescence. Therefore, progression to disease in schizophrenia-model mice can be prevented by treatments supporting vH-mPFC PV network function during a sensitive time window late in adolescence, suggesting therapeutic strategies to prevent the outbreak of schizophrenia.

Author Correction: Time units for learning involving maintenance of system-wide cFos expression in neuronal assemblies.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Author Correction: Infralimbic cortex is required for learning alternatives to prelimbic promoted associations through reciprocal connectivity.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Managing Neuronal Ensembles: Somatostatin Interneuron Subpopulations Shape and Protect Cortical Neuronal Ensembles for Learning.

Learning is accompanied by temporal compression and sharpening of neuronal firing sequences. In this issue of Neuron, Adler et al. (2019), using a motor skill paradigm and its variant, uncover a dual role for somatostatin interneuron regulation to support ensemble compaction and protection in learning.

Time units for learning involving maintenance of system-wide cFos expression in neuronal assemblies.

Repeated experiences may be integrated in succession during a learning process, or they may be combined as a whole within dedicated time windows to possibly promote quality control. Here we show that in Pavlovian, incremental and incidental learning, related information acquired within time windows of 5 h is combined to determine what mice learn. Trials required for learning had to occur within 5 h, when learning-related shared cues could produce association and interference. Upon acquisition, cFos expression was elevated during 5 h throughout specific system-wide neuronal assemblies. Time window function depended on network activity and cFos expression. Local cFos activity was required for distant assembly recruitment through network activity and distant BDNF. Activation of learning-related cFos assemblies was sufficient and necessary for time window function. Therefore, learning processes consist of dedicated 5 h time windows (time units for learning), involving maintenance of system-wide neuronal assemblies through network activity and cFos expression.

Author Correction: PV plasticity sustained through D1/5 dopamine signaling required for long-term memory consolidation.

In the version of this article initially published, the right panel in Fig. 2b was duplicated from the corresponding panel in Fig. 2c, and some data points in Fig. 3b were duplicated from Fig. 3a. None of the conclusions in the paper are affected. The errors have been corrected in the HTML and PDF versions of the article, and source data have been posted for the revised panels. The original and corrected figures are shown in the accompanying Author Correction.

Infralimbic cortex is required for learning alternatives to prelimbic promoted associations through reciprocal connectivity.

Prefrontal cortical areas mediate flexible adaptive control of behavior, but the specific contributions of individual areas and the circuit mechanisms through which they interact to modulate learning have remained poorly understood. Using viral tracing and pharmacogenetic techniques, we show that prelimbic (PreL) and infralimbic cortex (IL) exhibit reciprocal PreL↔IL layer 5/6 connectivity. In set-shifting tasks and in fear/extinction learning, activity in PreL is required during new learning to apply previously learned associations, whereas activity in IL is required to learn associations alternative to previous ones. IL→PreL connectivity is specifically required during IL-dependent learning, whereas reciprocal PreL↔IL connectivity is required during a time window of 12-14 h after association learning, to set up the role of IL in subsequent learning. Our results define specific and opposing roles of PreL and IL to together flexibly support new learning, and provide circuit evidence that IL-mediated learning of alternative associations depends on direct reciprocal PreL↔IL connectivity.

Parvalbumin Interneuron Plasticity for Consolidation of Reinforced Learning.

Parvalbumin (PV) basket cells are widespread local interneurons that inhibit principal neurons and each other through perisomatic boutons. They enhance network function and regulate local ensemble activities, particularly in the γ range. Organized network activity is critically important for long-term memory consolidation during a late time window 11-15 h after acquisition. Here, we discuss the role of learning-related plasticity in PV neurons for long-term memory consolidation. The plasticity can lead to enhanced (high-PV) or reduced (low-PV) expression of PV/GAD67. High-PV plasticity is induced upon definite reinforced learning in early-born PV basket cells, whereas low-PV plasticity is induced upon provisional reinforced learning in late-born PV basket cells. The plasticity is first detectable 6 h after acquisition, at the end of a time window for memory specification through experience, and is critically important 11-15 h after acquisition for enhanced network activity and long-term memory consolidation. High- and low-PV plasticity appear to regulate activity in distinct local networks of principal neurons and PV basket cells. These findings suggest how flexibility and stability in learning and memory might be implemented through parallel circuits and networks.

Functional and structural underpinnings of neuronal assembly formation in learning.

Learning and memory are associated with the formation and modification of neuronal assemblies: populations of neurons that encode what has been learned and mediate memory retrieval upon recall. Functional studies of neuronal assemblies have progressed dramatically thanks to recent technological advances. Here we discuss how a focus on assembly formation and consolidation has provided a powerful conceptual framework to relate mechanistic studies of synaptic and circuit plasticity to behaviorally relevant aspects of learning and memory. Neurons are likely recruited to particular learning-related assemblies as a function of their relative excitabilities and synaptic activation, followed by selective strengthening of pre-existing synapses, formation of new connections and elimination of outcompeted synapses to ensure memory formation. Mechanistically, these processes involve linking transcription to circuit modification. They include the expression of immediate early genes and specific molecular and cellular events, supported by network-wide activities that are shaped and modulated by local inhibitory microcircuits.

CLK2 inhibition ameliorates autistic features associated with SHANK3 deficiency.

SH3 and multiple ankyrin repeat domains 3 (SHANK3) haploinsufficiency is causative for the neurological features of Phelan-McDermid syndrome (PMDS), including a high risk of autism spectrum disorder (ASD). We used unbiased, quantitative proteomics to identify changes in the phosphoproteome of Shank3-deficient neurons. Down-regulation of protein kinase B (PKB/Akt)-mammalian target of rapamycin complex 1 (mTORC1) signaling resulted from enhanced phosphorylation and activation of serine/threonine protein phosphatase 2A (PP2A) regulatory subunit, B56ß, due to increased steady-state levels of its kinase, Cdc2-like kinase 2 (CLK2). Pharmacological and genetic activation of Akt or inhibition of CLK2 relieved synaptic deficits in Shank3-deficient and PMDS patient-derived neurons. CLK2 inhibition also restored normal sociability in a Shank3-deficient mouse model. Our study thereby provides a novel mechanistic and potentially therapeutic understanding of deregulated signaling downstream of Shank3 deficiency.

PV plasticity sustained through D1/5 dopamine signaling required for long-term memory consolidation.

Long-term consolidation of memories depends on processes occurring many hours after acquisition. Whether this involves plasticity that is specifically required for long-term consolidation remains unclear. We found that learning-induced plasticity of local parvalbumin (PV) basket cells was specifically required for long-term, but not short/intermediate-term, memory consolidation in mice. PV plasticity, which involves changes in PV and GAD67 expression and connectivity onto PV neurons, was regulated by cAMP signaling in PV neurons. Following induction, PV plasticity depended on local D1/5 dopamine receptor signaling at 0-5 h to regulate its magnitude, and at 12-14 h for its continuance, ensuring memory consolidation. D1/5 dopamine receptor activation selectively induced DARPP-32 and ERK phosphorylation in PV neurons. At 12-14 h, PV plasticity was required for enhanced sharp-wave ripple densities and c-Fos expression in pyramidal neurons. Our results reveal general network mechanisms of long-term memory consolidation that requires plasticity of PV basket cells induced after acquisition and sustained subsequently through D1/5 receptor signaling.

Inhibitory microcircuit modules in hippocampal learning.

It has recently become possible to investigate connectivities and roles of identified hippocampal GABAergic interneurons (INs) in behaving rodents. INs targeting distinct pyramidal neuron subcompartments are recruited dynamically at defined phases of behavior and learning. They include Parvalbumin Axo-axonic and perisomatic Basket cells, and Somatostatin radiatum-oriens and oriens-lacunosum moleculare cells. Each IN is in turn either activated or inhibited upon specific behavioral and network state requirements through specific inputs and neuromodulators. Subpopulations of these principal neurons and INs interconnect selectively, suggesting selective processing and routing of alternate information streams. First canonical functional modules have emerged, which will have to be further defined and linked to identified afferents and efferents towards a circuit understanding of how hippocampal networks support behavior.

Regulation of Parvalbumin Basket cell plasticity in rule learning.

Local inhibitory Parvalbumin (PV)-expressing Basket cell networks shift to one of two possible opposite configurations depending on whether behavioral learning involves acquisition of new information or consolidation of validated rules. This reflects the existence of PV Basket cell subpopulations with distinct schedules of neurogenesis, output target neurons and roles in learning. Plasticity of hippocampal early-born PV neurons is recruited in rule consolidation, whereas plasticity of late-born PV neurons is recruited in new information acquisition. This involves regulation of early-born PV neuron plasticity specifically through excitation, and of late-born PV neuron plasticity specifically through inhibition. Therefore, opposite learning requirements are implemented by distinct local networks involving PV Basket cell subpopulations specifically regulated through inhibition or excitation.

From intrinsic firing properties to selective neuronal vulnerability in neurodegenerative diseases.

Neurodegenerative diseases (NDDs) involve years of gradual preclinical progression. It is widely anticipated that in order to be effective, treatments should target early stages of disease, but we lack conceptual frameworks to identify and treat early manifestations relevant to disease progression. Here we discuss evidence that a focus on physiological features of neuronal subpopulations most vulnerable to NDDs, and how those features are affected in disease, points to signaling pathways controlling excitation in selectively vulnerable neurons, and to mechanisms regulating calcium and energy homeostasis. These hypotheses could be tested in neuronal stress tests involving animal models or patient-derived iPS cells.