Dynamics and Effective Connectivity in Bi- and Three–dimensional Neuronal Cultures: from Self–organization to Engineering

Abstract

This thesis was part of the European consortium MESOBRAIN, a team of 5 organizations that joined efforts in nanofabrication, cell culturing, imaging and data analysis to build tailored human 3D networks. The thesis timing was limited to 3 years, and several of the resources needed for its development were built from scratch. The main objective of this Ph.D. thesis was to explore complex characteristics of cortical neuronal cultures in terms of effective connectivity and exhaustive network analyses. This objective comprised four research lines: (i) The evaluation of neuronal network resilience and emerging plasticity mechanisms, (ii) the characterization of functional development to underline crucial timepoints in healthy neuronal networks, (iii) the study of 3D network interactions of neurons embedded inside an ECM--like environment, and (iv) the design, construction and viability inspection of neurons seeded on tiny 3D nanoprinted solid scaffold structures as a first step towards recreating cortical columns in vitro. For these multiple lines, we used either primary rat cultures (i,iii,iv) or human--derived neurons (ii). The former group corresponds to cultures with long established protocols that have been thoroughly studied in the field. The latter group corresponds to human neurons derived from iPSCs, a relatively novel model with promising and thrilling applications in regenerative medicine. Despite the increasing use of stem cells in neuroscience, complex systems and medicine, they still lack a thorough exploration in terms of neuronal and circuit formation as well as the properties of the emergent activity patterns. With either primary or stem cells, we explored the formation of neuronal circuits in 2D and 3D, characterized the effective connectivity and rendered a number of network traits. This Thesis combines experiments of highly difficult implementation with detailed data analysis. It was necessary to develop brand new protocols for culturing 3D neuronal networks and for human-derived neurons, the use of different microscopy setups the programming of object detection and tracking software and advance the analysis toolbox of calcium fluorescence data. First, resilience experiments on primary clustered neuronal cultures consisted on progressive perturbations through chemical receptor antagonists. This study represents an inspiring numerical--experimental model to comprehend the impact of plasticity mechanisms in the spontaneous activity of neuronal circuits. The results showed that, upon progressive connectivity blockade through chemical receptors' antagonists, only--excitatory neuronal networks displayed a surprising hyper--efficiency (HE) state for early--onset doses. As plasticity mechanisms influence the response of effective connectivity in the presence of perturbations, these compensatory mechanisms, usually disregarded, must be included in biological modeling as accurately as possible. Otherwise, episodes of functional rewiring and synaptic strengthening could mask important phenomena during experiments that alter channel communication. A simple algorithm that hypothesized an effective synaptic scaling was able to capture the hyper--efficiency state seen in experimental data, while percolation models wrongly predicted a progressive decay. The second research line was a sum of engineering efforts within the MESOBRAIN consortium, the European adventure to build 3D neuronal cultures embedded in hydrogels and with the presence of scaffolds. After several months of biomaterials testing, the candidate D--Clear resulted suitable for the construction of scaffolds, both with primary rat cells and hiPSCs, due to its good optical properties, manageability and biocompatibility. To our knowledge, D--Clear was never used before outside the orthodontic field and could provide a new catalogue of interesting designs for support and guidance of neuronal assemblies. Using this material, we developed a series of designs to offer support and guidance to cortical neurons in a 3D platform. The third research line focused on the study of neuronal development and cell-to-cell interactions in a semi-synthetic hydrogel that resembles the extracellular matrix of the brain. These hydrogel cultures keep the advantages of in vitro models while achieving an effective connectivity and architecture closer to in vivo. Finally, the fourth line of research applied cortical neurons from human-derived pluripotent stem cells to study key developmental stages and characterize the healthy maturation of these cells in vitro. As this technology has tremendous potential for regenerative medicine and to model neuronal diseases, it is urgent to consolidate the capacity of these human neuronal networks to reproduce efficient activity patterns of healthy patients, and explore the differences against the results obtained with animal models.

La presente tesis doctoral se enmarca en el contexto de la Física de la Materia Condensada, la Biofísica y la Neurociencia. Principalmente, se centra en el estudio de la conectividad funcional en cultivos neuronales bidimensionales (2D) y tridimensionales (3D). El trabajo se ha desarrollado en el Laboratorio del director de tesis Dr. Jordi Soriano, en la Facultad de Física de la Universitat de Barcelona, junto con el codirector Dr. Daniel Tornero, en el Hospital Clínic de Barcelona. Esta tesis forma parte del proyecto europeo MESO-BRAIN, del programa Future and Emergent Technologies (FET) de la Comisión Europea, Horizon2020. El trabajo de investigación combina experimentos con cultivos neuronales (de rata embrionaria o células humanas pluripotentes) y un análisis detallado en el contexto de teoría de redes y sistemas complejos. Los principales núcleos del trabajo realizado son los siguientes: (i) Actividad funcional en cultivos de redes neuronales y los mecanismos homeostáticos que emergen en presencia de perturbaciones; (ii) el desarrollo de herramientas de neuroingeniería para preparar cultivos ad hoc con conectividad dirigida mediante scaffolds; (iii) el análisis exhaustivo de los procesos de formación y madurez de redes funcionales humanas obtenidas de células madre pluripotentes inducidas, una nueva tecnología que promete revolucionar el campo de la medicina regenerativa; y (iv) la caracterización de cultivos neuronales 3D en estructuras que imitan la matriz extracelular natural de su entorno. Entre las diversas técnicas para la realización de cultivos tridimensionales, destacan los hidrogeles semi-sintéticos, constituidos en base a polímeros altamente hidratados con alta biocompatibilidad y cuyas propiedades mecánicas pueden ser manipuladas para obtener la estructura óptima según el tipo de tejido. En conjunto, los resultados de la presente tesis muestran la gran versatilidad de los cultivos neuronales y aportan avances relevantes en el estudio de plasticidad en redes neuronales, madurez y desarrollo tanto en 2D como en 3D, con sus correspondientes diferencias, incluyendo el uso de neuronas humanas derivadas de células madre inducidas. En el futuro, estos estudios nos permitirán incrementar nuestro conocimiento sobre el funcionamiento global del cerebro y avanzar en la investigación de diferentes enfermedades neurodegenerativas.