The dynamics of the nervous system are characterized by the intricate interactionsamong precisely arranged multiscale neural networks, arising from the innate capacity of the system for self-organization and adaptation to regulatory cues from the microenvironment. However, the fundamental principles governing organization and function of neural networks remain largely unknown. In recent years, in vitro engineered interfaces have emerged as a viable approach to study neural networks at the mesoscale, providing a highly controlled microenvironment and allowing to mimic key aspects of in vivo neural networks, while minimizing the associated complexities by adopting a reductionist approach. In this M.Sc thesis, we used novel models for in vitro neural culturing to recapitulate significant aspects from in vivo topology and manipulate the structure-function dynamics within populations of neurons. To attain modular organization of segregated neuronal populations we used novel multi-nodal microfluidic devices, featuring afferent-efferent connectivity promoted via Tesla-valve inspired microchannels. The presence of structural modularity and unidirectionality was validated by employing viral tools for cellular delivery of genes for expressing distinct fluorescent proteins within the segregated neural population in two-nodal microfluidics. To capture their functionaldynamics, we implemented and optimized a system for optogenetics and calcium imaging. Furthermore, we demonstrate that altering the number of microchannels within a five-nodal microfluidic coupled with microelectrode array can influence inter-nodal connectivity. We then selectively perturbed one node within the five-nodal microfluidic and showed notable alternations in functional connectivity among the nodes, as well as in firing rates within each node. Finally, we tested the biocompatibility of novel 3D microfabricated interfaces with neuronal cultures, aiming to facilitate the formation of tridimensional neural networks while retaining the simplicity of handling neural cultures. This work highlights the immense potential of advanced neuroengineered in vitro models to recapitulate microscale topological organization of neural networks and investigate their structure-function dynamics in healthy and perturbed conditions.