Deposition chamber technology as building blocks for a standardized brain-on-chip framework

Neurological disorders like Alzheimer's and Parkinson's disease significantly impact the aging population, imposing substantial healthcare costs. Progress in diagnosis and treatment is hindered by the brain's complexity and the limitations of current *in vivo* models for preclinical trials, which lack structural and functional translationality. Organ-on-a-chip (OoC) microfluidic technology offers potential by regulating connectivity and directionality between neuronal populations. However, current OoC technologies lack control over the complete architecture of neural circuits, particularly the internal structure of individual nodes. This study introduces a method to standardize microfluidic strategies for reconstructing complex, compartmentalized neuronal networks, paving the way for investigating structure-function relationships in neurological diseases. The authors aim to develop a method for efficiently scaling and designing neuro-engineered microfluidic devices to control homogeneous neuron seeding with a targeted number of cells within each node. This technology will allow for the creation of multinode neurofluidic chips, holding structural connectivity patterns between several nodes. A specific application would be modeling the basal ganglia loop affected in Parkinson's disease, potentially creating a minimalist yet relevant model of the disease when coupled with human-derived iPSC neurons.

Existing in vitro models utilizing microfluidic technology have made strides in regulating connectivity and directionality between compartmentalized neuronal populations, creating simplified neuronal networks. Neuro-engineered OoC microfluidics offer control over cellular environments, enabling co-culture of different neural cell types while maintaining fluid isolation. These approaches serve as minimalistic in vitro models of in vivo neural circuits. However, while the focus has been on connectivity patterns between nodes, less attention has been paid to the internal architecture of these nodes. Current techniques, whether microfluidic, colloidal, or scaffolded, lack full control over the neural circuit architecture, leaving a critical gap in the physiological relevance of brain-on-chip approaches. The standardization of microfluidic strategies to reconstruct complex interconnected neuronal networks is therefore necessary.

The authors propose using "Three-Dimensional Deposition Chambers" (DCs) as supplemental compartments integrated into existing microfluidic designs. These DCs are positioned on an upper level to prevent neuronal connections between the DCs and the inlet/outlet channels. A key innovation is the use of channels with distinct dimensions (inlet wider and higher than outlet) to control fluid velocity. Hydrostatic pressure differences between inlet and outlet reservoirs generate controlled flow, eliminating the need for pumps. The geometry and proportions of the DCs are defined to control the number of neurons seeded, correlating chamber surface size with the quantity of neurons. The optimal flow velocity (Vch) within the DC is determined based on the settling velocity of neurons (Vsedi) using the equation Vch - Vsedi * Hch / Lch = 1, where Hch and Lch are the height and length of the DC. The hydrodynamic resistance is calculated considering major head losses due to viscous effects (using equations to account for channel width, height, length, flow rate, fluid density and viscosity). Minor head losses are neglected. To optimize, channel widths and heights are fixed, and the lengths of the inlet and outlet channels are adjusted to achieve the desired flow velocity. Reservoir dimensions and infused cell suspension volumes are also determined. A model of the flow profile and deposition rate of cells was developed to monitor flow speed, simplifying the multiphase process with three assumptions: laminar flow (Poiseuille flow), flow velocity dependence on hydrostatic pressure, hydraulic resistance and fluid viscosity, and a unicellular layer of cells in a regular hexagonal close packing configuration. The neuronal coverage of the DC surface is calculated based on neuron number, neuron radius and chamber surface size. The change in neuronal coverage over time is determined by the neuron concentration and the volume of the suspension entering the DC.

The study demonstrates a method for controlling the homogeneous seeding of neurons with a targeted number of cells within individual nodes of a microfluidic device. The approach is efficient across a wide range of neuronal quantities, maintaining uniform surface coverage. The use of hydrostatic pressure to generate flow eliminates the need for pumps. The authors present equations and methodologies for calculating optimal flow velocity, hydrodynamic resistance, and neuronal deposition rate. Experimental results (e.g. Figure 1 and 2) show successful fluid renewal and homogeneous neuronal deposition within the designed deposition chambers. The described methodology allows for precise control over cell density and placement in microfluidic chambers.

This novel microfluidic design significantly advances the creation of brain-on-a-chip models by addressing the limitation of controlling cellular density and spatial organization within individual nodes. The ability to precisely control neuron numbers within each chamber, while maintaining homogeneous distribution, is critical for building more physiologically relevant *in vitro* models of neural circuits. The elimination of pumping systems simplifies the device operation and potentially increases accessibility. This methodology is applicable to various neural populations, expanding the potential for modeling diverse neural circuits and investigating disease mechanisms. The described technology represents a significant step towards standardization in brain-on-a-chip technology, facilitating broader adoption and reproducibility in neuroscience research. The work specifically highlights potential for modeling Parkinson's disease by creating a microfluidic architecture representing the basal ganglia loop.

This study presents a significant advancement in brain-on-a-chip technology by introducing a novel deposition chamber design enabling precise and uniform seeding of neurons. The method's efficiency across a range of cell quantities and elimination of pumps simplifies the process. Future work could focus on integrating this technology with other established techniques for controlling connectivity between nodes to create more complex and physiologically accurate brain models and validating the method with human-derived iPSC neurons to model specific neurodegenerative diseases.

The model simplifies the complex multiphase process with assumptions regarding laminar flow, fluid properties, and cell packing. Further validation and refinement of the model may be needed for different cell types or experimental conditions. The current design focuses on two connected chambers; scaling to more complex, multi-node systems should be investigated. The experimental validation mainly uses rat hippocampal neurons, so further work is necessary to test with diverse neuronal cell types.