Imaging Lipid Droplets by Electron Microscopy - Methods in Cell Biology - page 227

13.1.3 Results and considerations
13.1.3.1 Observation of small LDs
LDs smaller than 100 nm in diameter are often difficult to identify by conventional
EM due to the medium electron density of the core and the absence of the bilayer
membrane. Imidazole-buffered osmium tetroxide was shown to increase the electron
density of LDs ( Angermuller & Fahimi, 1982 ), but preservation of other cellular
structures is often compromised using this method.
The LD containing DHA-rich TG shows an extremely high electron density and
can be distinguished clearly from the surrounding cytoplasmic matrix. By treating
cells with 0.1 mM DHA for 30 min, small electron-dense LDs in the diameter range
of 50-75 nm can be observed ( Fig. 13.2 B). Because LDs are depleted by incubation
in the lipid-deficient medium, they are thought to be nascent LDs that were made of
newly synthesized TG. In defining the region where new LDs form, it is important
that this method can make very small LDs observable.
13.1.3.2 Incorporation of newly synthesized TG into existing LDs
The method described here is based on the result that the electron density of TG
formed by different fatty acids is correlated with the number of unsaturated bonds.
The electron density of LDs is highest in cells cultured with DHA (22:6; 22 carbon, 6
unsaturated bonds), followed by those with linoleic acid (18:2) and OA (18:1)
( Cheng et al., 2009 ). By measuring with the scale using internal standards
( Fig. 13.2 C), the difference in the electron density of LDs can be objectively mea-
sured. Importantly, a normal variation of other factors, such as section thickness and
strength of lead staining, does not affect the relative electron density significantly.
When 3Y1 cells pretreated with OA for 12 h are cultured with DHA, the average
electron density of LDs increases in a time-dependent manner ( Fig. 13.2 D). Al-
though the electron density of LDs varies at any one time point, LDs showing the
electron density as high as that of DHA-only LDs (see Fig. 13.2 B) cannot be ob-
served until 2 h after the DHA addition ( Fig. 13.2 E). When the order of OA and
DHA treatments is reversed (i.e., first DHA for 12 h and second OA for 0-4 h),
the relative electron density of LDs decreases gradually and the variation among
LDs is very small again. The results indicate that newly synthesized TG is largely
incorporated into existing TG-rich LDs and does not form LDs on its own in 3Y1
cells. It is of note that the result is different in other cell types including white
adipocytes.
In comparison with the above results on TG-rich LDs, CE-rich LDs are different
in that they show heterogeneity in the pattern of incorporating DHA-rich TG ( Cheng
et al., 2009 ). That is, some LDs show an increased electron density homogenously,
whereas others reveal high and low electron-dense areas in an LD. The latter pattern
probably occurs because existing CEs and DHA-derived TGs do not mix well and are
segregated from each other. Nonetheless, the result indicates that newly synthesized
TG is incorporated into existing LDs whether those LDs predominantly contain
TG or CE.