Abstract
Purpose: :
Neuroprotectin D1 (NPD1) attenuates laser-induced CNV and induces activated microglia redistribution (Sheets et al., ARVO 2011). As microglia morphology correlates with activation, we quantitatively assessed microglia morphology to further elucidate the effects of NPD1 signaling on activated microglia in experimental CNV. In addition to the classic 2D method, we measured the full 3D shape of microglia. To our knowledge, this is the first 3D quantitative analysis of microglia morphology.
Methods: :
Three laser CNV lesions were made per eye. Right eyes received 16 ng NPD1 topically, once daily for seven days; left eyes were vehicle. Choroid flatmounts were prepared at 15 days post-laser, labeled with isolectin B4, CD11b, and Hoechst, and visualized by confocal microscopy. 3D image stacks of microglia, peripheral to lesions, were segmented, binarized and the parameters volume, surface area, area, perimeter, form factor, solidity, convexity, and fractal dimension were measured in 3D and 2D maximum intensity projections using ImageJ. 3D measures were obtained with our custom plugin, Convex Hull 3D. Data were iteratively Gaussianized (Goerg, GM 2011) and assessed by anova.
Results: :
NPD1 increased microglia cell volume 120% from 549 μm3 in controls to 1209 μm3 (p<0.001) and surface area 96% from 2304 to 4520 μm2 (p<0.001). Much of this change occurred antiparallel to the z-axis, as maximum intensity projections also revealed significant increases in analogous measures; area increased from 544 to 1012 μm2 (p<0.001) and perimeter increased from 376 to 668 μm (p=0.001). NPD1 reduced form factor 50% (3D, p=0.026) and 64% (2D, p=0.009) and increased fractal dimension 45% (3D, p=0.005) and 13% (2D, p<0.001). NPD1 reduced 2D, but not 3D, convexity 29% (p<0.001). Solidity was not significantly different between treatments.
Conclusions: :
3D morphologic analysis is tractable and superior to 2D, which generally overestimates shape differences. Form factor reveals control microglia were more spherical and rounded than NPD1 microglia. Fractal dimension indicates rougher microglial surfaces with NPD1 over control. Our results imply that activated microglia are more amoeboid-like in CNV and that NPD1 promotes larger, more ramified microglia. Combined with our previous results of CNV inhibition, the present study suggests NPD1 may attenuate CNV, in part, by regulating microglia activation. This new target of NPD1 signaling is being explored as an early intervention in age-related macular degeneration.
Keywords: microglia • choroid: neovascularization • age-related macular degeneration