Deformation of lipid droplets in fixed samples

Nile red, Sudan III, and oil red O have been used to stain lipid droplets (LDs) for fluorescence microscopy. We noticed that LDs labeled by Nile red are different in appearance from those stained by the latter two dyes. To understand the cause of the difference, we used sequential labeling procedures (first LD stain–photography–quenching–second LD stain–photography), and examined the effect of several factors. Immunofluorescence labeling for adipose differentiation-related protein (ADRP), an LD marker, was also observed comparatively with the lipid stains. As a result, we found that ethanol and isopropanol used for Sudan III and oil red O staining, respectively, and glycerol used for mounting, cause fusion of adjacent LDs even in glutaraldehyde-fixed samples. By the same treatment, immunofluorescence labeling for ADRP was dislocated to the rim of large LDs that were formed as a result of the artifactual fusion. The result indicates that the LD structure can be better observed with Nile red than with Sudan III or oil red O.