Prefrontal-hippocampal dysfunction is considered as main cause of cognitive deficits in mental disorders, yet its ontogeny and mechanisms throughout life are still largely unknown. Using mouse models that mimic the etiology and symptoms of schizophrenia, we showed during the first two funding periods that, shortly after birth, the long-range coupling between prefrontal cortex and hippocampus is reduced due to structural and functional deficits of specific neuronal populations. Preliminary data identified juvenile age as an additional period of high vulnerability for disease. Here, we will elucidate the cellular substrate of juvenile prefrontal-hippocampal miswiring leading to poorer cognitive performance and identify electrophysiological markers of dysfunction common to mice and prodromal humans.

We use optogenetic and electrophysiological approaches to investigate how specific neuronal populations encode spatial memories. By activity-dependent expression of optogenetic silencers, we are able to shut off mossy cells that were highly active during encoding of a specific spatial location and test the effects on water maze performance. As we could show, learning of a new location typically re-activates dentate gyrus neurons encoding for the old location. We will investigate the importance of this reactivation for successful reversal learning and monitor learning-induced connectivity changes between the left and right hemisphere.

How does the brain know what it has to remember and what it may safely forget? Dopamine is a central neuromodulator signaling salience and novelty and it is important for learning and memory formation. In this project we will investigate how dopamine controls network activity and functional connectivity in the dorsal hippocampus of mice during different behavioral states. Using two-photon imaging in the hippocampus during behavior in combination with optogenetic and chemogenetic manipulation of locus coeruleus and ventral tegmental area we aim at unravelling the specific roles of these long-range projecting areas for defined behavioral states and hippocampus-dependent learning and memory-formation.

This project aims at elucidating the impact of stress and major stress mediators on memory reconsolidation and the underlying brain networks. We predict that stress and glucocorticoid activation impair reconsolidation by interfering with the recruitment of and crosstalk in limbic, prefrontal and large-scale frontoparietal memory networks, whereas noradrenergic stimulation may have the opposite effect. In order to test these predictions, we will combine experimental stress-induction and pharmacological elevation of glucocorticoid and noradrenergic activity with funtional MRI recordings of memory network dynamics during encoding, reactivation and test.

The present study will test two hypotheses in human visual and multisensory neural networks: (1) The development of top-down communication lags behind the development of bottom-up processing; (2) the development of top-down processing depends to a larger extent on early sensory experience than the development of bottom-up processing. Two EEG paradigms, a visual attention and a multisensory paradigm, will be developed in human adults. They will consecutively be adapted to children (5-11 year old, prospective developmental approach) and sight-recovery individuals (retrospective developmental approach) to test these two hypotheses, respectively.

The network mechanisms involved in the spatial integration of tactile information into saccade planning and in the required coordinate transformations from skin-centered to eye-centered reference frames are investigated using EEG-based connectivity measures. We hypothesize that spatial integration relies on interareal coupling and not on local processing. We test this by perturbing the identified network using repetitive TMS. EEG and TMS results will be modeled on a mesoscopic scale based on a subgraph of the human connectome.

In the absence of major anatomical deficits, abnormal information processing and disturbed multi-site communication between brain regions seem to represent an important pathophysiological mechanism underlying neurodevelopmental disorders like schizophrenia. During the first funding period, we showed that the functional communication within prefrontal-hippocampal networks is disturbed already during the early development when genetic and environmental factors converge. Using optogenetics and electrophysiology the present proposal aims at identifying the cellular elements underlying the early network dysfunction.

We will use optogenetic intervention during formation and recall of spatial memories to investigate the generation of cognitive maps and the overlap between engrams of specific spatial positions. Using activity-dependent expression of light-gated chloride channels, we will determine how multiple positions in a single environment are encoded. We will investigate how entorhinal cortex, dentate gyrus and CA1 interact when spatial memories are retrieved, updated during spatial reversal learning, and consolidated during sleep.

Sensory deprivation is accompanied by a crossmodal activation of deprived sensory brain regions. For ex-ample, in the blind occipital cortical activity has been suggested to result in compensatory performance changes. This project uses an extensive training paradigm with different types of stimuli. It is hypothesized that blind individuals reach a higher asymptotic performance level than sighted controls which in turn is re-lated to a training induced integration of parts of occipital cortex into the processing network seen in sighted controls. Thus, this project tests whether extensive changes in cortico-cortical interactions rather than a com-pensatory allocation of functions to individual deprived brain areas is the critical mechanism of crossmodal plasticity. 

The project is designed to investigate the neuronal activities and multi-site connectivity in entorhinal cortex and hippocampus that control information transfer between these regions. To this aim, we will analyze the physiological and behavioral consequences caused by the alteration of intrinsic resonance properties of neu-rons. Specifically, we will analyze how transgenic inhibition of Kv7/M and HCN/h currents in specific neu-ronal populations in mice affects extracellular current oscillations and unit activity in the entorhinal cortex and hippocampus at different stages of learning and memory consolidation and retrieval.

This project aims at investigating the synaptic multi-site interactions between the hippocampus and the neo-cortex throughout the process of memory consolidation. So far it is not known if cortical and hippocampal synaptic plasticity consolidates in a similar or dissimilar manner and if continuous re-consolidation in the hippocampus is essential for long-term memory even when cortical representations already exist. Moreover, nothing is known about the underlying molecular and cellular mechanisms. The basis for the project is the finding that the activity-regulated gene Arc/Arg3.1 is essential for the consolidation of long-term synaptic plasticity and remote memory. In genetically engineered mice we will restrict deletion of Arc/Arg3.1 either to the hippocampus or to the cortex, and will employ behavioural tests to assess the development of remote memory. Electrophysiological recordings will be conducted in vivo and in vitro to measure changes in synap-tic transmission within the cortico-hippocampal circuitry.

This project investigates mechanisms of integration and segregation of multimodal sensory and sensorimotor events. We use unsupervised learning to optimize properties of static representations of audio-visual stimuli captured in different reference frames. Next, interactions on slow, medium and fast time scales are included within a unified formalism. We generalize these results to structured hierarchical networks modelled accord-ing to the statistical constraints of the cortical connectome. Finally, we apply these results to learning and plasticity after focal cortical lesions. This project directly relates to several other projects of this SFB pro-posal and links descriptive and mechanistic levels and thereby contributes to our understanding of cortical function.