hPSC-derived microglia shape neuronal morphology and enhance network activity in vitro 161 5 Since the reduced total axonal and dendritic segment numbers point towards altered neuronal morphology, we did additional post-hoc testing to determine if there are specific morphological changes. On the axonal level we showed that in cultures with microglia-like cells each neuron had on average less segments, that were longer, with less extremities and reduced nodes (Fig. 2G). Interestingly for dendrites we only observe the average length of the dendritic segments per neuron to be increased in the presence of microglia-like cells (Fig. 2H, p = 0.027, Bonferroni corrected n = 8), while per neuron the number of segments, extremities and nodes were not significantly altered. To study synaptic pruning, we measured and observed that both the absolute number of synapses (Mean -MG: 8140, Mean +MG: 6365, p = 0.0002) as well as the synapse density were significantly reduced (p = 0.018, Fig. 2I) in the presence of microglia-like cells. In conclusion, addition of hPSC-derived microglia-like cells significantly reduced nuclear debris and affects neuronal morphology. Addition of microglia alters neuronal network activity To investigate whether the addition of microglia-like cells affects different features of neuronal network activity as well, we analyzed in parallel to the morphological analysis, the same neuron and microglia batches on multi-electrode arrays (MEA) (Fig. 3A). During 6 weeks of microglia-neuron co-culturing, neuronal networks with microglia-like cells showed an increase in activity which was reflected by a significant increase in active electrodes (Figure 3B-C, p = 0.007). Additional post-hoc testing showed that changes in activity are caused by increased spike rate (Fig. 3D, p = 0.0298, Bonferroni corrected n = 13, Supplementary Table IV). The significant effects are observed independent of neuronal background (Supplementary Fig. 4A-H). Other parameters such as burst and network burst counts show a similar trend of increased activity in the presence of microglia-like cells, however, these do not hold after Bonferroni correction (Fig. 3E-F, Supplementary Table IV). We also observed a trend of increased spiking within bursts (Supplementary Fig. 5A) and network bursts (Supplementary Fig. 5B) when neurons were in co-culture with microglialike cells. Additionally, the shape of bursts and network bursts as defined by duration (Supplementary Fig. 5C-D) and spike count (Supplementary Fig. 5E-F) might show a trend of increased duration and increased spike count. For frequency there only is a trend towards higher spike frequency in network bursts (Supplementary Fig. 5G-H, Supplementary Table IV). Together, we show a significant increase in activity as well as trends in other activity parameters that also point towards increased activity and higher synchronicity in cortical neuron networks when co-cultured with hPSC-derived microglia-like cells.
RkJQdWJsaXNoZXIy MTk4NDMw