PhD Projects

The goal of this project is to understand the mechanisms and the operations performed in the cerebellum and cerebello-thalamo-prefrontal pathway during fear learning in mice. We have preliminary data showing anatomical and physiological connections between cerebellum, and prefrontal cortex via medio-dorsal thalamus (MD), which is known to be important for fear extinction. We will now evaluate the correlated patterns of neuronal activity with fear learning using electrophysiological recording in vivo, during different phases of fear conditioning in the limbic system (MD thalamus, prefrontal cortex) in freely moving animals coupled with optogenetic stimulation (theta-burst) to test the cerebello-limbic functional link and to modulate this link in order to modify emotional behaviour.

Anxiety will be measured in open field experiments, in dark/light compartment boxes and elevated cross labyrinth, before and after optogenetic cerebellar modulation. A startle test will also be performed using our accelerometer to accurately quantify the startle responses. The feasibility of this project is very high because all techniques and analysis are already implemented in the respective labs. Collaboration with Essen (ESR13) will compare the effects of cerebellar theta burst stimulation in mice with similar non -invasive stimulation in human subjects.

Overall, these experiments will provide new insights into the neurophysiological processes (cell discharge, oscillations, synchrony across limbic structures) controlled by the cerebellum during fear learning and extinction and will determine the effect of cerebellar manipulations on these processes.

This project is hosted by École Normale Supérieure, Paris, France.

This project will use the latest deep-brain calcium imaging technology (miniscope, Inscopix), to monitor activity of identified cell types within the cerebellum and brainstem of mice, with a focus on defensive states. This project will address the hypothesis that cerebellar-PAG circuits form a crucial network element that modulates aversive emotions via somatosensory integration of behavioural and autonomic responses. To characterize specific functional pathways, e.g. ascending vs. descending cerebellar-PAG connections, we will use double-conditional, intersectional viral strategies to target cell-types with distinct projection targets (see ESR6). Freely moving mice will be subjected to behavioural assays such as the open field test and elevated plus maze to evoke different levels of fear and/or anxiety and adaptive behavioural coping strategies. Mice will also be equipped with subcutaneously implanted electrodes for recording of muscle activity and electrocardiogram. Furthermore, we will employ the recently developed conditioned flight paradigm to investigate cue- and context- dependent switching between passive and active threat coping. This will allow monitoring of potential “online prediction errors” and their putative relation with moment-to-moment integrated defense states. The results will be compared to the electrophysiological findings obtained in ESRs1, 2 and 5 and directly compared to calcium imaging results in ESR4. Collaboration with Pavia in analysis of calcium imaging data to incorporate into a computational model.

The resulting comprehensive dataset will allow to identify specific neuronal correlates within cerebellar-PAG circuits for distinct integrated defense states. Furthermore, temporal decoding of calcium activity will reveal the neural population dynamics of “online prediction error” generation.

This project is hosted by University Hospital Würzburg, Germany.

The latest in vivo calcium imaging technology (miniscope – INSCOPIX) will be used to monitor neuronal population activity in different cerebellar cortical modules during the acquisition and extinction of emotional behaviour in freely moving mice. The tasks will be designed in order track cerebellar activity in relation to error prediction. Cerebellar neuronal activity will be recorded during baseline, acquisition and extinction phases of conditioned fear behaviour. During the tasks, we will track mouse behavioural responses and correlate them with different stimuli and environmental cues. The same experiments will then be the basis for simulations by ESR9. Collaboration with Wurzburg and INSCOPIX.

This project will determine the extent of cerebellar activation during emotional behaviour and the contribution of different modules to encoding prediction error.

This project is hosted by Universitá Pavia Italy.

This project will explore the possibility that cerebellar modules differ in their control of affective behaviour, depending on anxiety phenotype. The details of the project will be tailored to the student’s interest but will include comparison in high versus low anxiety phenotypes of rat and manipulation of neural circuits using the latest viral transfection methods (e.g. cre-dependent DREADDs). The effect of circuit manipulations in the cerebellum will be studied during behavioural paradigms that elicit negative versus positive affective responses (e.g. behaviours that elicit fear/anxiety such as fear conditioning, elevated plus maze, open field and lever press for reward).

There will be a focus on integrated behavioural outputs including somatomotor, autonomic and ultrasonic vocalisations. The experiments may include the use of state-of-the-art multi brain site electrophysiological recording and/or calcium imaging, together with ECG to investigate neural encoding in different anxiety phenotypes (including the possibility of gender differences); and the effect of the manipulations on neural circuit and cardiovascular activity.

These experiments will provide new insights into cerebellar contributions to integrated behavioural responses elicited by affective state.

The results will be compared to the electrophysiological and calcium imaging findings obtained in ESRs 1-4 and related clinical research (ESR10). Collaboration with Wurzburg will provide guidance on the use of viral techniques and calcium imaging (e.g. Purkinje cell specific promoters).

This project is hosted by University of Bristol, UK.

To characterize the neuronal circuit architecture for information flow between the cerebellum and limbic system, we will anatomically map the neural circuits that link modules of the cerebellar cortex via different cerebellar nuclei (CN), to the PAG, amygdala and prefrontal cortex in mice. We will combine trans-synaptic rabies virus tracing and adeno-associated viral vectors with antero- and retrograde tracing capabilities to investigate the projection patterns of selected PAG, amygdala and prefrontal cortex cell types targeting cerebellar vermis and CN. Importantly, this will allow us to also characterize putative collaterals to other brain regions, such as cardiorespiratory centres in the brainstem. We will also identify the neuronal cell types that project from the CN to different cellular targets within PAG, amygdala and prefrontal cortex.

In a second step, and cross-informed by results obtained through calcium imaging approaches (e.g. by ESR3-4), we will address the functional significance of the identified cerebellar-limbic pathways. We will use viral strategies as described above to express excitatory or inhibitory optical actuators in selected circuit elements. Freely moving mice implanted with optical fibres as well as ECG/EMG electrodes will undergo behavioural assays for fear and anxiety, such as elevated plus maze, open field and conditioned flight paradigm. During different integrated defensive states (freezing/flight, brady-/tachycardia) and various stages of cued and contextual fear conditioning, we will interfere with cerebellar information flow, i.e. prediction error generation to manipulate defensive action selection upon threat exposure as well as fear learning. Anatomical results will be compared to cFos mapping in ESR7. Collaboration with Paris to compare anatomical with electrophysiological results.

Viral vector-mediated intersectional anatomical tracing strategies will provide new insights into cerebellar-brainstem connectivity with cell-type specificity. Temporally precise manipulation of cerebellar inputs to defined limbic neuronal subpopulations (and vice versa) will elucidate causal relationships of neuronal activity and prediction error generation as well as behavioural and autonomic defensive responses during fear and anxiety states.

This project is hosted by University Hospital Würzburg, Germany.

Chronic pain is a multidimensional experience including potentially debilitating affective responses and changes in emotional responses processing. There is a link between mood disorders (such as anxiety and depression) and individual susceptibility to the development of chronic pain (e.g. Woo, Rev pain, 2010; McWIlliams et al, Pain, 2003). Phenotypes of anxiety in humans and rodents range from high to low and this project seeks to understand the link between this and those individuals more susceptible to developing chronic pain.

It is hypothesised that chronic pain could be a dysfunction of extinction learning i.e. an inability to forget pain causing a persistent fear of pain, anxiety and pain catastrophisation.

We (and others- Jimena et al; Utz et al) have previously shown how cerebellar vermal regions are involved in fear extinction learning. This project aims to further understand how the cerebellum contributes to emotional control via extinction learning using animal models in a neuropathic pain state and human studies.

This project is hosted by University of Bristol, UK.

This project will determine how signal processing differs between different parts of the same cerebellar module, in particular between vermal lobules IV-V, VI-VII, VIII, and IX-X. High density multielectrode arrays (HD-MEA – 3-Brain) will be used to characterize spontaneous activity from Purkinje cells, Golgi cells and molecular layer interneurons within each lobule, as well as responses to electrical stimulation of mossy fibre bundles. Advanced techniques for LFP modelling (Diwakar et al, 2011; Casali et al., 2019) will be used to extract information about the spatio-temporal organization of the underlying neuronal circuit activity. The HD-MEA work will be complemented by patch-clamp recordings to provide evidence of the cellular mechanisms unveiled by HD-MEA. Collaboration with Bristol and 3-Brain.

This project will provide new insights into neuronal activity within cerebellar modules and how the signal processing varies between different lobules.

This project is hosted by Universitá Pavia Italy.

An advanced integration of the data obtained in this ITN with literature will be achieved by computational and robotic modeling. In the framework of the Human Brain Project, of which this host institution is part, a cerebellar model is developed that includes detailed microcircuit connectivity and neuronal properties (Casali et al., 2019; Geminiani et al., 2018, 2019a, 2019b). This background will be taken as the starting point of this project, where the modules corresponding to the cerebellar vermis in relation to emotional control will be investigated. The model will be adapted and extended in order to account for the data obtained by the other participants to this project (e.g. ESR8 for MEA recordings, ESR5-7 for cerebellar structural and functional connectivity) and will be exploited to simulate the data obtained on fear conditioning learning tasks (e.g. ESR1-4). Collaboration with all beneficiaries.

This project will address the concept of a generalized cerebellar algorithm and to determine potential differences in the computation performed in different vermal lobules. Moreover, it will address the role of cerebellum in fear conditioning learning task.

This project is hosted by Universitá Pavia Italy.

The human cerebellum is well known for its contribution to motor learning, and disorders in motor learning have been related to motor performance deficits in patients with cerebellar disease. The cerebellum is likely also involved in learning and memory processes in the cognitive and emotional domains, but this has been studied in much less detail. Our overarching aim is to get a fuller understanding of the contribution of the human cerebellum to learning and memory of fear, an important emotion for survival. To provide further evidence that the human cerebellum contributes to the different phases  of fear learning we will perform fear conditioning studies in healthy human participants in the 7T MRI scanner including pharmacological interventions. ESR10 and ESR13 will work closely together.

Young and healthy participants will perform fear conditioning tasks in the 7T MRI scanner. In part of the experiments dopaminergic and anti-dopaminergic drugs will be administered. To determine possible sex differences male and female participants will be tested. Differential fear conditioning will be performed using a 3-day design. fMRI analysis will be performed during the presentation, the prediction, and the unexpected omission of aversive signals. Fear conditioning paradigms will be harmonised between the three fMRI research groups involved in the ITN to aid comparison: Uppsala, Edinburgh and Essen. Resting state data will be acquired at the beginning of all fMRI experiments at each site to test whether strength of functional connectivity in emotional networks predicts behavioural outcome. Furthermore, MRI data analysis will be performed in close collaboration between the three 7T sites, and will include univariate and multivariate methods on reactivity/connectivity. Our group has performed structural and functional 7T MRI studies in healthy participants for many years with a focus on the cerebellum. A fear conditioning set-up in the 7T scanner is available. The findings will be compared with fMRI data in rodents (ESR11), and in patients with ataxias (ESR13) and anxiety disorders (ESR15). In collaboration with Bristol (ESR5) fMRI imaging and electrophysiological results will be compared in the medial and lateral cerebellum during fear conditioning in humans and rodents, respectively.

Experiments will provide further evidence that the human cerebellum contributes to learning in the emotional domain. We expect that different parts of the cerebellar cortex will be involved in fear learning, including more medial and lateral areas. More specifically, we expect that the unexpected omission of aversive signals, which is rewarding, will result in increased cerebellar activation reflecting processing of prediction errors. fMRI signals are expected to be modified by pharmacological interventions.

This project is hosted by University Hospital Essen, Germany.

This project will characterise fear related cerebellar circuit properties using a range of behavioural, anatomical and electrophysiology/manipulation techniques in a rodent model of ASD.

A growing body of literature suggests that many behavioural and pathological changes in ASD can be associated with cerebellar circuit dysfunction. Indeed, multiple rodent models of ASD display cerebellar and behavioural abnormalities characteristic of impairments identified in individuals with autism, including fear and anxiety. Furthermore, unpublished work from our group has revealed robust fear phenotypes in transgenic rat models of ASD.

Given that prediction errors, encoded in cerebellar circuits, are thought to be essential to the fear extinction process and likely underlie many core ASD phenotypes, such deficits may thus reflect aberrant connectivity/activity within distributed cerebellar-fear circuits.

Therefore, this project will map fear related circuit cerebellar circuit activations (e.g. via cFOS) in a transgenic rat model of ASD coupled with in vitro/vivo electrophysiology to record interactions between specific cerebellar regions and emotion networks across a fear conditioning paradigm. Finally, manipulation of specific cerebellar output pathways may be used in an attempt to rescue fear behavioural deficits. The results will be compared and contrasted with those obtained in ESRs 1, 6 and 7. Collaboration with Paris on circuit manipulation experiments.

These experiments will determine how distributed cerebellar-fear circuits contribute to emotional processing in a rat model of ASD.

This project is hosted by University of Edinburgh, UK.

This project will use 9.4T fMRI in a rat model of ASD to understand the contribution of the cerebellum to disordered emotional control.

Given prediction errors are encoded in cerebellar circuits are thought to be essential to the fear extinction process and are thought to underlie many core ASD symptoms, such deficits may thus reflect aberrant activity within distributed cerebellar-fear circuits. However, despite the extensive behavioural and pathological changes in ASD that can be associated with the cerebellum, surprisingly little is known about the processing of predication errors within this structure during, for example, associative learning in pre-clinical models of the disorder.

MRI has the potential to provide translatable and non-invasive system-wide and repeatable measures of brain structure (volume measurement) and function (fMRI). Moreover, this approach enables findings from basic neuroscience in rodent models to be translated into clinical studies. fMRI of rodents is established within the Simons Initiative for the Developing Brain (making this project highly feasible) and enables assessment of temporal and spatial changes in network activation patterns in response to specific tasks (e.g. associative memory learning tasks).

Aim 1 of this project will combine targeted silencing of cerebellar Purkinje cells and fMRI in male and female rats that model ASD and wild-type controls to chart functional connections between the cerebellum and fear circuitry at rest. Aim 2 will use fMRI to correlate cerebellar activation with components of the fear circuitry (e.g. PAG, prefrontal cortex, amygdala) related to recall and extinction of conditioned fear. This set of experiments will provide the first insights into distributed cerebellar activity in rodent models of ASD linked to associative learning. This project will be conducted in close collaboration with Uppsala and Essen. This will allow us to standardise our behavioural protocols, data acquisition and analysis. The findings will be compared with electrophysiological data in rodents (ESRs 1, 6, 9), and 7T fMRI data in patients with cerebellar (ESR13) and anxiety disorders (ESR15). Collaboration with Essen and Uppsala on 7T fMRI.

These experiments will determine how distributed cerebellar-fear circuits contribute to emotional processing in a rat model of ASD.

This project is hosted by University of Edinburgh, UK.

The human cerebellum is well known for its contribution to motor learning, and disorders in motor learning have been related to motor performance deficits in patients with cerebellar disease. The cerebellum is likely also involved in learning and memory processes in the cognitive and emotional domains, but this has been studied in much less detail. Our overarching aim is to get a fuller understanding of the contribution of the human cerebellum to learning and memory of fear, an important emotion for survival. To provide further evidence that the human cerebellum contributes to fear learning we will perform fear conditioning studies in patients with cerebellar degeneration in the 7T MRI scanner. ESR10 and ESR13 will work closely together.

Patients with cerebellar cortical degeneration and age- and sex-matched controls will be tested in a 7T Magnetom Terra MR scanner. Differential fear conditioning will be performed using a 3-day design. fMRI activity will be assessed in the CS-US time window (to assess fMRI activity related to prediction of the US), and, equally important, at the time the US is expected but does not occur (in interspersed CS-only and initial extinction trials; to assess prediction error). fMRI analysis will include functional connectivity analysis between the cerebellum and other brain areas involved in fear conditioning. Behavioural studies will be complemented by testing an approach-avoidance paradigm in patients with cerebellar degeneration in close collaboration with the group in Uppsala. The host group has extensive experience of structural and functional 7T MR imaging in patients with cerebellar degeneration and healthy participants, and has a well-established ataxia clinic for patient recruitment. A fear conditioning set-up in the 7T scanner is available. Fear conditioning paradigms will be harmonised between the three fMRI research groups involved in the ITN to aid comparison: Uppsala, Edinburgh and Essen. Resting state data will be acquired at the beginning of all fMRI experiments at each site to test whether strength of functional connectivity in emotional networks predicts behavioural outcome. Furthermore, MRI data analysis will be performed in close collaboration between the three 7T sites, and will include univariate and multivariate methods on reactivity/connectivity. The findings will be compared with fMRI data in rodents (ESR11), and in patients with ataxias (ESR13) and anxiety disorders (ESR15).

Experiments will provide further evidence that the human cerebellum contributes to learning in the emotional domain. Experiments will allow a fuller understanding of the interactions between the cerebellum and the known fear acquisition and extinction networks. Furthermore, close collaboration between the three MRI sites in the ITN will allow a comprehensive understanding of changes in emotional networks not only in cerebellar disease, but also neurodevelopmental and anxiety disorders, an important prerequisite to develop future treatments.

This project is hosted by University Hospital Essen, Germany.

This project will use functional magnetic resonance imaging (fMRI) to understand the contribution of the cerebellum to altered affective processing in patients with anxiety disorders and depression. Animal studies show that manipulations of cerebellar functioning affect fear and anxiety-related behavior and neuroimaging trials in humans support cerebellar involvement in anxiety and mood disorders. However, essential knowledge regarding the function of the cerebellum in the aetiology and treatment of these disorders is lacking.

Patients with affective disorders and healthy controls will be recruited for participation in fMRI trials. We will compare differential neural activation and functional connectivity patterns of the cerebellum during experimental emotional challenges in comparison to non-emotional stimuli. Multiple fMRI experimental designs will be used, including a novel method on moment-to-moment variability of the neural signal. Moreover, structural integrity (e.g., voxel-based morphometry) of the cerebellum will be compared across patients and healthy controls and related to functional interconnectivity between cerebellum and affect-processing brain areas. We will also evaluate whether emotion-related neural responsivity of the cerebellum predicts symptom severity and response to treatment with selective serotonin reuptake inhibitors (SSRIs) and/or cognitive behavioral therapy (CBT). Neuroimaging and treatment data will be acquired from ongoing studies, our large data base on social anxiety disorder and, if feasible, additional data made available through our ongoing international research collaborations on anxiety and depression.

The findings will be compared with imaging data in rodent models of ASD (ESR11), and humans with cerebellar disease (ESR13). Data analysis will be performed in collaboration with Pavia (ESR9) to develop computational models of neural networks in emotional disorders.

Results of the proposed experiments will further clarify the role of the cerebellum in promoting susceptibility to disorders of the affective spectrum. In patients, altered activation and connectivity could be expected in a cerebellum – limbic – frontocortical circuit, that in turn may be related to symptom severity and predictive of treatment outcome.

This project is hosted by Uppsala Universitet/Karolinska Institutet, Sweden.

This project will employ neuroimaging, psychophysiological measures, psychological experimental tasks, and pharmacological challenge to disentangle the involvement of the human cerebellum in fear conditioning and emotional processing and decision-making. The cerebellum has vast afferent and efferent projections to traditional fear and reward areas such as the amygdala, periaqueductal gray, insula, anterior cingulate cortex, medial prefrontal cortex, and striatum, and is thus well-situated to influence emotional processes. However, the role of cerebellum in these processes is not fully elucidated.

Age-related differences in fear conditioning processes have been noted, including impaired fear extinction in adolescence, and the suggested neural mechanisms have focused on immature prefrontal cortex-amygdala connectivity. Although the cerebellum is involved in fear conditioning, little is known of its role in the development of fear conditioning processes. Here, we will utilize a cross-sectional sample of children (6-9 years), adolescents (13-17 years) and adults (30-40 years) that have undergone structural 3T MRI (T1 weighted images) and a 2-day Pavlovian fear conditioning task using skin conductance and pupil dilation as indices of fear response. Developmental differences in cerebellum morphology and structural connectivity/covariance with other brain regions, such as the amygdala and anterior cingulate cortex, will be related to fear acquisition, extinction and return of fear.

Further, the role of the cerebellum in caffeine-induced anxiety will be studied using 7T fMRI. Caffeine is the most consumed psychoactive drug globally, and, whereas it has many beneficial effects, it also has anxiogenic and panicogenic effects. Certain genetic polymorphisms have been linked to increased anxiogenic effects of caffeine in behavioral studies, but the neural mechanisms of these effects have not been studied. Here, we will focus on how the cerebellum modulates the acute effects of a single dose caffeine on processing of emotional stimuli and emotional decision-making during approach-avoidance conflicts. Carriers and non-carriers of risk polymorphisms will be compared to further elucidate the neural underpinnings of caffeine-induced anxiety.

Findings will be related to fear extinction learning in rodents (ESR7), 7T fMRI data in rodents (ESR11) and in healthy participants and patients with cerebellar disease (ESR10, ESR13). Collaboration with Essen, Edinburgh on fMRI data analysis, and with Bristol in fear extinction learning in rodents.

Results of the experiments  will elucidate the role of the cerebellum at different neurodevelopmental stages in fear conditioning and its involvement in emotional processing during an acute caffeine challenge. Expected results include cerebellar modulation of caffeine-induced effects on emotional processing and decision-making during approach-avoidance conflicts, including connectivity with amygdala-striatal-frontocortical networks.

This project is hosted by Uppsala Universitet, Sweden.