Breaking the Burst: Unveiling Mechanisms Behind Fragmented Network Bursts in Patient-derived Neurons

Abstract

Fragmented network bursts (NBs) are observed as a phenotypic driver in many patient-derived neuronal networks on multi-electrode arrays (MEAs), but the pathophysiological mechanisms underlying this phenomenon are unknown. Here, we used our previously developed biophysically detailedin silicomodel to investigate these mechanisms. Fragmentation of NBs in our model simulations occurred only when the level of short-term synaptic depression (STD) was enhanced, suggesting that STD is a key player. Experimental validation with Dynasore, an STD enhancer, induced fragmented NBs in healthy neuronal networksin vitro. Additionally, we showed that strong asynchronous neurotransmitter release, NMDA currents, or short-term facilitation (STF) can support the emergence of multiple fragments in NBs by producing excitation that persists after high-frequency firing stops. Our results provide important insights into disease mechanisms and potential pharmaceutical targets for neurological disorders modeled using hiPSC-derived neurons.

Automated inference of disease mechanisms in patient-hiPSC-derived neuronal networks

“…Additionally, SBI forecasted a significant increase in STD strength in both GEFS+ and DS, along with a larger STD time constant for DS and increased asynchronous release in GEFS+. These predictions align with previous findings implicating these mechanisms in the observed phenotypes of GEFS+ and DS [11]. Furthermore, lower sEPSC amplitudes and frequencies were observed in DS neurons, possibly indicating more depressed synapses [10].…”

“…Less intuitively, we found a strong positive correlation between the NMDA conductance and the strength of short-term synaptic depression (U (STD)). Literature suggests that both mechanisms strongly but inversely influence the NB duration (NBD) MEA feature [17,11]. To investigate this, we computed the sensitivity of the NBD MEA feature to the model parameters and indeed found NBD to be highly sensitive to both NMDA conductance and the strength of STD (Supplemental Figure 2E), suggesting these parameters can compensate for each other.…”

“…Previous in vitro experiments revealed lower sodium current densities, altered action potential (AP) decay dynamics and firing patterns suggestive of a reduction in potassium currents, and reduced spontaneous excitatory post-synaptic current (sEPSC) frequency and amplitudes in GEFS+ and DS networks compared to Control [7, 10]. Moreover, increased STD was recently suggested to be important in GEFS+ and DS networks to explain the phenotype [11]. We found that the posterior distributions inferred with SBI visually differed between healthy control networks and DS and GEFS+ (Figure 4A).…”

“…These data often exhibit distinct characteristics that differ between healthy and diseased networks, reflecting underlying pathological molecular mechanisms [2, 3, 5, 6, 7, 21]. While some mechanisms are explicitly linked to specific activity patterns [17, 11], characterizing them definitively through quantitative analysis of the data remains challenging. Additional experiments to nonetheless uncover these mechanisms, such as patch-clamp measurements or RNA-sequencing, are costly and time-consuming and may not cover all possible molecular pathways, potentially overlooking crucial mechanisms [10, 21].…”

“…Computational models are invaluable in bridging the gap between experimental observations and the (patho-)physiological mechanisms underlying them [9]. Previously, we have developed a bio-physically detailed computational model of hiPSC-derived excitatory neuronal networks on MEA [10, 11], that can dissect the effect of specific cellular changes on network activity. The parameters of this in silico model describe key physiological characteristics of the neurons, the synapses, and the network.…”