Long-term morphological and functional dynamics of human stem cell-derived neuronal networks on high-density micro-electrode arrays

Abstract

Comprehensive electrophysiological characterizations of human induced pluripotent stem cell (hiPSC)-derived neuronal networks are essential to determine to what extent these in vitro models recapitulate the functional features of in vivo neuronal circuits. High-density micro-electrode arrays (HD-MEAs) offer non-invasive recording with the best spatial and temporal resolution possible to date. For 3 months, we tracked the morphology and activity features of developing networks derived from a transgenic hiPSC line in which neurogenesis is inducible by neurogenic transcription factor overexpression. Our morphological data revealed large-scale structural changes from homogeneously distributed neurons in the first month to the formation of neuronal clusters over time. This led to a constant shift in position of neuronal cells and clusters on HD-MEAs and corresponding changes in spatial distribution of the network activity maps. Network activity appeared as scarce action potentials (APs), evolved as local bursts with longer duration and changed to network-wide synchronized bursts with higher frequencies but shorter duration over time, resembling the emerging burst features found in the developing human brain. Instantaneous firing rate data indicated that the fraction of fast spiking neurons (150–600 Hz) increases sharply after 63 days post induction (dpi). Inhibition of glutamatergic synapses erased burst features from network activity profiles and confirmed the presence of mature excitatory neurotransmission. The application of GABAergic receptor antagonists profoundly changed the bursting profile of the network at 120 dpi. This indicated a GABAergic switch from excitatory to inhibitory neurotransmission during circuit development and maturation. Our results suggested that an emerging GABAergic system at older culture ages is involved in regulating spontaneous network bursts. In conclusion, our data showed that long-term and continuous microscopy and electrophysiology readouts are crucial for a meaningful characterization of morphological and functional maturation in stem cell-derived human networks. Most importantly, assessing the level and duration of functional maturation is key to subject these human neuronal circuits on HD-MEAs for basic and biomedical applications.

Human neural network activity reacts to gravity changes in vitro

“…iNGN-derived neurons require an additional time for functional maturation ( Lam et al, 2017 ; Schmieder et al, 2022 ). During this maturation phase a connected network of neurons develops, network bursts arise, and a GABAergic system evolves, resembling human brain development ( Habibey et al, 2022 ). In the first 4 days of induction, an increase in the expression of glutamatergic genes was measured ( Busskamp et al, 2014 ).…”

“…At 30 days post induction (dpi), about 2.3% of neurons have been shown to be GABAergic (Lu et al, 2019). This fraction increases over time and recent results suggest an increase in the number of inhibitory neurons after 60 dpi (Habibey et al, 2022). More detailed functional characterization of long-term development of iNGN networks has shown that the majority of iNGN neurons have functional synapses after 21 days and functional AMPA/KA receptors are present in iNGN neurons as confirmed by immunohistochemistry and patch-clamp experiments (Lam et al, 2017).…”

Microengineered 2D and 3D modular neuronal networks represent structure-function relationship

“…The in vitro network function is mainly studied multi-electrode arrays (MEAs) electrophysiology and optical tools like calcium imaging and optogenetics [11,14,15,22,23]. Non-invasive in vitro electrophysiology tools for multi-site long-term recordings, including transparent MEAs and high-density MEAs, are commercially available [13,22,[24][25][26][27]. They have been applied for mapping the functional connectivity patterns in randomly connected 2D neuronal networks [14,26,28,29].…”

“…However, random networks do not recapitulate the modular structure of the brain circuits [28]. In addition, in vitro network structure undertakes changes by culture age due to the shifts in the position of the neuronal cell body and axonal branches [14,21,27]. It means that a specific electrode does not necessarily record from same neurons or part of the circuit over weeks and months of network development [27].…”

“…In addition, in vitro network structure undertakes changes by culture age due to the shifts in the position of the neuronal cell body and axonal branches [14,21,27]. It means that a specific electrode does not necessarily record from same neurons or part of the circuit over weeks and months of network development [27]. This can lead to inconsistency in tracking activity and functional features, particularly for long-term experiments [27].…”