DPhil Student University of Oxford, United Kingdom
Induced pluripotent stem cell (iPSC) technology has revolutionized our capacity to experiment directly on live human cells. However, understanding of neuronal (patho)physiology remains hampered by over-simplified iPSC-derived monocultures which often fail to capture the physiologically relevant synaptic connections unique and critical to neuronal development and functions. Leveraging on recent advancements of the microfluidic technology including polydimethylsiloxane-based open multi-compartment or novel free-style fluid-walled microfluidics we established in vitro models of neuronal circuits using iPSC-derived neurons. On our custom microfluidic platforms, various iPSC-derived neuron subtypes were cultured and oriented in a physiologically relevant manner. Network of large neuronal soma was cultured exclusively within large pre-designed chambers that allows little cell mobility to the outside due to either physical barricade or relative pressure differences. Meanwhile, neurites were guided following a combination of both pressure difference and chemotatic gradients to project extensively outwards, forming targeted connections with different neuronal populations. Custom direction of neuronal connections was also facilitated by conduits of 1 mm long which consequently allowed complete passing of only elongated axons but not short stubby dendrites. As a result, these setups recapitulated both the compartmentalisation of cell soma from their neurites and the endogenous collaterals outgrowth onto other neurons of interest. Virtually any neuronal circuit with custom orientation could be established in vitro in this manner given the modular nature of the microfluidic-based neural circuit approach. We established the proof-of-principle the in vitro model of the human presynaptic striatum microcircuit on open microfluidic, capturing the endogenous excitatory and modulatory inputs from cortical and midbrain dopaminergic neurons, respectively, onto striatal medium spiny neurons. Human neurons within this original three-component neuronal circuit model displayed significantly improved functional maturation as compared to those cultured in the circuit-absent environment. Interestingly, in the context of familial Parkinson’s Disease neurons cultured in the microcircuit exhibited emerging pathological properties which might be a novel therapeutic consideration. We also demonstrated the feasibility of modelling neuronal circuit employing the novel fluid-walled microfluidics. The enhanced accessibility and reconfigurability unique of the fluid walls allow precise damage to any component of the neuronal circuit and efficient re-establishment of the circuit for subsequent monitoring of ensuing events, respectively. Hence, directed axotomy of the human corticostriatal pathway in vitro revealed the important role of target-derived factors in facilitating axonal regeneration. Our results highlight the advantage and necessity of various microfluidic platforms in modelling human neuronal circuits in a low cost and high throughput manner. The plethora of different microfluidic modules with distinctive practical strengths enriches the current tool box of iPSC-derived modelling for various experimental and screening paradigms. Such advanced modelling of human neurons recapitulating more aspects of the complex microenvironment of the human brain will permit emerging physiological properties that could be otherwise elusive.
