Figure 2. Proinflammatory cytokines mediate a redox-dependent reorganization of the neuronal actin cytoskeleton.

Serum-starved SH-SY5Y human neuroblastoma cells grown on collagen were incubated either with 10 μM DPI (an NADPH oxidoreductase inhibitor), or 40 μM MnTBAP (a catalytic superoxide dismutase mimetic) prior to acute cytokine exposure (200 ng/ml, 15 min). Following paraformaldehyde fixation, cultures were stained with rhodamine phalloidin and actin filament organization visualized by confocal microscopy. (A) Pretreatment with DPI (upper panel) or MnTBAP (lower panel) prevented transient rearrangement of the actin cyoskeleton of SH-SY5Y cells upon cytokine exposure. SH-SY5Y displayed a condensed morphology in the presence of DPI (upper panel, left). However, DPI pretreatment of SH-SY5Y cells negated lamellipodia formation, membrane ruffling, and cell spreading upon exposure to TNFα or Il-1β, respectively (upper panel, middle and right). SH-SY5Y cells also retained an atrophic morphology following preincubation with MnTBAP (lower panel, left). MnTBAP largely suppressed lamellipodia formation and membrane ruffling (arrowheads) upon addition of TNFα or Il-1β. Note, pseudopodia-like structures were observed in the presence of cytokines particularly in MnBAP-pretreated cultures. (Scale bar = 10 μm). (B) Quantitative analysis (see criterion in Experimental Methods) revealed that DPI-pretreatment suppressed lamellipodia formation and membrane ruffling upon acute exposure to TNFα or Il-1β (grey bars) compared to an absence of DPI (black bars, *p<0.001) to levels indistinguishable from controls (open bars). (C) Pretreatment with MnTBAP greatly reduced lamellipodia formation and membrane ruffling upon addition of cytokines (grey bars, **p<0.01 compared to control) as opposed to an absence of MnTBAP (black bars, *p<0.01). All values were normalized to controls and represent the mean of at least three independent experiments ± standard deviations n=180).