Fig. 4: Networks of neuronal cells on soft PDMS surfaces.

Visual examination of fluorescence images of neuronal cells suggests that surface stiffness can influence cell-clustering (a). We used image-analysis algorithms and networks-science to examine quantitatively the topological characteristics of cell-networks on the substrates. Fluorescence images of cells were gray-scale converted (b) and processed to extract cell-centers (c). Then, cell-centers were linked using the Waxman algorithm and a density-based rule (d). The small world coefficient (SW, a topological measure of networks) of neuronal-cell graphs as a function of surface elasticity, determined \(24\,h\) from culture – the diagram suggests that the ability of cells to form structured networks decreases with \(E\) (e). The small world coefficient of neuronal-cell graphs as a function of time for a fixed value of the Young’s modulus \(E=2.65{MPa}\) (f). Correlation between the SW coefficient of neuronal cell networks forming on a substrate and the substrate elasticity, for different values of culture time: \(24\), \(48\), \(72\), \(96\,h\) (g). Correlation between the SW coefficient and time, for different values of substrate elasticity: \(0.55\), \(1\), \(1.88\), \(2.65{MPa}\) (h). Data in Fig. 4e–h are represented by mean ± standard deviation (sample size\(\sim\)50 for each data point).