Chapter 5 174 areas bigger than 10 µm2 using the Find Nuclei module. Any areas between 10 and 50 µm2 with a roundness lower than <0.8 were considered to be nuclear debris. Mature neuron number was determined by determining how many nuclei (DAPI area > 50 µm2) had a high intensity of NeuN. First, we excluded nuclei over 240 µm2 in size, as neurons are known to have small nuclei. Next, the Select population module was used and for each timepoint intensity cut-offs for NeuN positivity were manually set by inspection of the settings on at least 5 pictures from different wells. Before proceeding, Select Population module on HuNu intensity was used to determine if neurons had indeed human origin. After this quality control, Find Neurites module with the build in CSIRO neurite analysis 2 was used to identify NF+ neurites. This analysis provides an overview of axonal morphological properties corrected for the amount of neurons present. Similarly, MAP2 positive nuclei were identified and the dendrite morphology obtained. Next, using the Select Region module the dendrites were investigated on containing synaptophysin positive puncta using the Find Spots module. Again, intensity parameters for detection of spots were manually set for each time point based on at least 5 pictures from different wells. The amount of synaptophysin positive puncta is corrected per unit of dendritic length. Statistical analysis with vs without microglia-like cells To determine whether microglia numbers and morphology were affected by neuronal background or changed over time, a two-way ANOVA using the cell line and timepoints as independent variables was performed. P-values were corrected for multiple testing using Bonferroni correction (n = 18). Four proxy metrics were used to determine if the co-culture morphology is changed upon microglia addition: nuclear debris, mature neuron count (NeuN+) and number of axonal and dendritic segments. The conditions with and without microglia-like cells are compared per proxy metric. No directional hypothesis was specified and all comparisons were performed through two-sided tests. Tests were performed using a linear model including neuronal background and timepoint as covariates. A Shapiro-Wilk test for normality was performed on the residuals, and when normality was rejected (p < 0.05), a Mann-Whitney U test was performed instead. To assess whether there was an overall impact on neuronal cultures, while minimizing the number of multiple comparisons, p-values for each of the four proxy metrics were aggregated using Fisher’s method. If Fisher’s method was significant, post-hoc analysis was performed using the outcomes of the tests on the proxy metrics. Dendrite and axon morphology post-hoc analysis were performed using either Ordinary Least Square
RkJQdWJsaXNoZXIy MTk4NDMw