Structural architecture of the Neuronal-Glial-Vascular system

Brain functionality relies on the neuronal-glial-vascular (NGV) ensemble for energy support. However, the details of the complex biological mechanisms involved in these processes and the microscopic interactions between these three components are not yet fully understood. Astrocytes, an essential component of the NGV ensemble, extend ramified processes to the vasculature and to neuronal synapses to provide neuronal energy support, homeostatic control, and multi-directional signaling with the vasculature, neurons, and neighboring astrocytes. As key links between the components of the NGV circuitry, the astrocytes are essential for drug delivery in the brain and are involved in the progression of neurodegenerative diseases. Therefore, comprehensive models of the neuronal-glial-vascular (NGV) ensemble are essential for understanding the role of astrocytes in the formation and function of the complex networks within the brain and its pathological conditions. A complete computational model of this ensemble has not yet been developed.

This thesis presents the first model of a data-driven digital reconstruction of the neuronal-glial-vascular (NGV) ensemble at a micrometer anatomical resolution. Combining the sparse literature data and the few available detailed biological reconstructions, I have computationally generated the structural architecture of a neocortical NGV circuit that forms a functional column of the juvenile rat neocortex and consists of neurons, protoplasmic astrocytes, and the microvasculature, including their pairwise connectivities. This data-driven approach allows for incremental refinement as more experimental data become available, new biophysical models get published, and new questions arise. The NGV circuit is validated against a plethora of literature sources to ensure its biological fidelity: it successfully reproduces the spatial organization of the astrocytes, their morphological characteristics as well as their volume occupancy, and the overlap with their neighbors.

The power of the computational model of NGV lies in its ability to serve as a framework for addressing long-standing questions that cannot be experimentally investigated due to the complexity of the microscopic systems and the limitations of current techniques to observe all components simultaneously. In this thesis, I have performed experiments to investigate why astrocytes acquire their particular shapes, the organizational principles of NGV that lead to the observed biological network architectures, and the effect of astrocytic density on endfoot organization. The circuit's structural analysis showed that astrocytes optimize their positions and spacing from each other to provide the vasculature with a uniform coverage for trophic support and signaling. By increasing the density of astrocytes in NGV circuits, I discovered a limit in astrocyte's ability to make perivascular projections because of the vascular spatial occupancy, which constrains the extent of astrocyte morphology. Thus, their role in linking vasculature to neurons constrains their organization via the proportional relationship of density and microdomain shrinkage due to contact spacing.

By addressing these questions, I demonstrated how this model can serve as an exploratory tool that provides a window into the complexity of the NGV architecture.