The nervous system is the most complex product of biological evolution and its functioning is based on the connections of complex systems. The constantly changing patterns of activity of its billions of interactive units represent the fundamental physical basis of human behavior and experience. For these reasons, the study of the nervous system is complex both from an anatomic and a functional point of view. A multiscale approach is therefore necessary to combine experimental findings from different levels of complexity, such as the micro- and macroscale, and to bridge the gap between them by investigating how the mesoscale interacts with the others. One of the most complex circuits which turns out to be investigated is the cerebro-cerebellar circuit. Thus, the aim of this thesis was the anatomo-functional characterization of this circuit by developing a multiscale approach since none of the previous investigations focusing on this circuit have combined data from different scales of complexity. To develop this multiscale approach, I performed several steps. First of all, starting from literature, I created schemes for the connections involving deep cerebellar nuclei, either considering them as a whole or as a single nucleus separately. Then, mesoscale data were extracted from the Allen Mouse Brain Connectivity Atlas database, which consists of 428 anterograde tracing experiments in wild type C57BL/6J mice obtained by mapping fluorescently labeled neuronal projections brain-wide. This has provided a computational neuroinformatic platform able to investigate axonal connectivity of the mouse brain. Macroscale data were instead elaborated from 10 subjects included in the Human Connectome Project (HCP) database. Tractography, exploiting diffusion magnetic resonance imaging, is the only non-invasive method able to reconstruct axonal bundles in human in vivo. In details, a constrained spherical deconvolution algorithm was combined with probabilistic tractography to reconstruct the whole-brain tractogram of each subject. Then, to select a specific axonal bundle it is fundamental to constrain bundle extraction by defining a priori regions of interest (ROIs). It is worth noting that I based ROIs selection uniquely on schemes that I previously drawn up from literature, instead of iteratively correct results on the basis of empirical reconstructions. Connectome matrices in mice and humans were extracted at meso- and macroscale level, respectively. At mesoscale level, the connectome extracted from the Allen Mouse Brain Connectivity Atlas database demonstrated that injection cerebellar sites presented contralateral connections with the thalamus and the cerebral cortex while ipsilateral connections remained at the level of the brainstem and cerebellum, in agreement with the anatomical decussations of the tracts passing through the cerebellar peduncles. Furthermore, cerebellar lobules and nuclei showed many connections with multiple subthalamic nuclei, but less than half with the cerebral cortex suggesting that the pathways investigated with viral tracers are anterograde and mainly monosynaptic. The connectome derived from tractography at macroscale level confirmed a greater efferent cerebellar connectivity with the thalamus instead of the cerebral cortex, while afferent cerebellar connectivity was better identified using tractography instead of viral tracing. In conclusion, the present multiscale approach resulted to be effective to compare and merge results across species, e.g. mouse and human, for validating results at different levels of complexity and reducing the gap between meso- and macroscale experiments. Therefore, future investigations are warranted to further quantify the cerebro-cerebellar connectivity and to validate other brain circuits across species, with the final goal of investigating the key role that evolution plays in the development or conservation of patterns and connections at the level of the nervous system.
Il sistema nervoso è il prodotto più complesso dell'evoluzione biologica e il suo funzionamento si basa sulle connessioni di sistemi complessi. Lo studio del sistema nervoso è complesso sia da un punto di vista anatomico che funzionale.Un approccio multiscala è necessario per combinare i risultati sperimentali provenienti da diversi livelli di complessità, come la micro e la macroscala,e per colmare il divario tra loro indagando come la mesoscala interagisce con le altre. Uno dei circuiti più complessi risulta essere il circuito cerebro-cerebellare. L'obiettivo di questa tesi era la caratterizzazione anatomo-funzionale di questo circuito sviluppando un approccio multiscala poiché nessuna delle indagini precedenti su questo circuito ha combinato dati da diverse scale di complessità. Per sviluppare questo approccio multiscala,ho eseguito diversi passaggi.Partendo dalla letteratura,ho creato degli schemi per le connessioni che coinvolgono i nuclei cerebellari profondi,sia considerandoli nel loro insieme che separatamente come un singolo nucleo. I dati su mesoscala sono stati estratti dal database Allen Mouse Brain Connectivity Atlas, che consiste in 428 esperimenti di tracciatura anterograda in topi C57BL/6J wild type ottenuti mappando proiezioni neuronali marcate con fluorescenza in tutto il cervello, fornendo una piattaforma neuroinformatica computazionale in grado di indagare la connettività assonale del cervello di topo. I dati su macroscala sono stati invece elaborati da 10 soggetti inclusi nel database dello Human Connectome Project (HCP). La trattografia, sfruttando la dMRI, è l'unico metodo non invasivo in grado di ricostruire i fasci assonali nell'uomo in vivo. In dettaglio, un algoritmo di deconvoluzione sferica vincolata è stato combinato con la trattografia probabilistica per ricostruire il trattogramma del cervello intero di ogni soggetto. Per selezionare un fascio assonale specifico è fondamentale,per vincolare l'estrazione del fascio, definire a priori regioni di interesse (ROI). Vale la pena sottolineare che ho basato la selezione delle ROIs unicamente su schemi che ho precedentemente elaborato dalla letteratura, invece di correggere iterativamente i risultati sulla base di ricostruzioni empiriche. Le matrici del connettoma nei topi e negli esseri umani sono state estratte rispettivamente a livello di meso- e macroscala. A livello di mesoscala,il connettoma estratto dal database Allen Mouse Brain Connectivity Atlas ha dimostrato che i siti cerebellari di iniezione presentavano connessioni controlaterali con il talamo e la corteccia cerebrale, mentre le connessioni ipsilaterali rimanevano a livello del tronco encefalico e del cervelletto, in accordo con le decussazioni anatomiche dei tratti che attraversano i peduncoli cerebellari. Inoltre, i lobuli e i nuclei cerebellari hanno mostrato molte connessioni con più nuclei subtalamici, ma meno della metà con la corteccia cerebrale, suggerendo che i percorsi indagati con traccianti virali sono anterogradi e principalmente monosinaptici.Il connettoma derivato dalla trattografia a livello di macroscala ha confermato una maggiore connettività cerebellare efferente con il talamo invece della corteccia cerebrale, mentre la connettività cerebellare afferente è stata meglio identificata usando la trattografia invece che il tracciato virale. In conclusione,l'attuale approccio multiscala è risultato efficace per confrontare e unire i risultati tra le specie, ad esempio il topo e l'uomo, per convalidare i risultati a diversi livelli di complessità e ridurre il divario tra esperimenti di meso- e macroscala.Pertanto, le indagini future sono garantite per quantificare ulteriormente la connettività cerebro-cerebellare e per convalidare altri circuiti cerebrali attraverso le specie, con l'obiettivo finale di indagare il ruolo chiave che l'evoluzione gioca nello sviluppo o conservazione di modelli e connessioni a livello del sistema nervoso.
Il circuito cerebro-cerebellare studiato a diverse scale di complessità: dalla connettività micro a quella macroscala
VASIRANI, GIULIA
2020/2021
Abstract
The nervous system is the most complex product of biological evolution and its functioning is based on the connections of complex systems. The constantly changing patterns of activity of its billions of interactive units represent the fundamental physical basis of human behavior and experience. For these reasons, the study of the nervous system is complex both from an anatomic and a functional point of view. A multiscale approach is therefore necessary to combine experimental findings from different levels of complexity, such as the micro- and macroscale, and to bridge the gap between them by investigating how the mesoscale interacts with the others. One of the most complex circuits which turns out to be investigated is the cerebro-cerebellar circuit. Thus, the aim of this thesis was the anatomo-functional characterization of this circuit by developing a multiscale approach since none of the previous investigations focusing on this circuit have combined data from different scales of complexity. To develop this multiscale approach, I performed several steps. First of all, starting from literature, I created schemes for the connections involving deep cerebellar nuclei, either considering them as a whole or as a single nucleus separately. Then, mesoscale data were extracted from the Allen Mouse Brain Connectivity Atlas database, which consists of 428 anterograde tracing experiments in wild type C57BL/6J mice obtained by mapping fluorescently labeled neuronal projections brain-wide. This has provided a computational neuroinformatic platform able to investigate axonal connectivity of the mouse brain. Macroscale data were instead elaborated from 10 subjects included in the Human Connectome Project (HCP) database. Tractography, exploiting diffusion magnetic resonance imaging, is the only non-invasive method able to reconstruct axonal bundles in human in vivo. In details, a constrained spherical deconvolution algorithm was combined with probabilistic tractography to reconstruct the whole-brain tractogram of each subject. Then, to select a specific axonal bundle it is fundamental to constrain bundle extraction by defining a priori regions of interest (ROIs). It is worth noting that I based ROIs selection uniquely on schemes that I previously drawn up from literature, instead of iteratively correct results on the basis of empirical reconstructions. Connectome matrices in mice and humans were extracted at meso- and macroscale level, respectively. At mesoscale level, the connectome extracted from the Allen Mouse Brain Connectivity Atlas database demonstrated that injection cerebellar sites presented contralateral connections with the thalamus and the cerebral cortex while ipsilateral connections remained at the level of the brainstem and cerebellum, in agreement with the anatomical decussations of the tracts passing through the cerebellar peduncles. Furthermore, cerebellar lobules and nuclei showed many connections with multiple subthalamic nuclei, but less than half with the cerebral cortex suggesting that the pathways investigated with viral tracers are anterograde and mainly monosynaptic. The connectome derived from tractography at macroscale level confirmed a greater efferent cerebellar connectivity with the thalamus instead of the cerebral cortex, while afferent cerebellar connectivity was better identified using tractography instead of viral tracing. In conclusion, the present multiscale approach resulted to be effective to compare and merge results across species, e.g. mouse and human, for validating results at different levels of complexity and reducing the gap between meso- and macroscale experiments. Therefore, future investigations are warranted to further quantify the cerebro-cerebellar connectivity and to validate other brain circuits across species, with the final goal of investigating the key role that evolution plays in the development or conservation of patterns and connections at the level of the nervous system.È consentito all'utente scaricare e condividere i documenti disponibili a testo pieno in UNITESI UNIPV nel rispetto della licenza Creative Commons del tipo CC BY NC ND.
Per contatti: unitesi@unipv.it
https://hdl.handle.net/20.500.14239/14111