“Array tomography” is a method used to observe the fine structure of cells and tissues in a three-dimensional view. In this method, serial ultrathin sections in the ribbon state (ribbons) are mounted on a solid substrate and observed by scanning electron microscopy (SEM). The method may also be used in conjunction with post-embedding immunocytochemistry. However, it is difficult to mount many serial ribbons on a substrate manually. We developed an inexpensive laboratory-made device that mounts ribbons by pulling a nylon fishing line and lifting the substrate up from the water in a knife boat. Using this device, we succeeded in mounting several ribbons consisting a mean of 205.6 (SD: 37.7) serial ultrathin sections on 1.25 (SD: 0.06) × 1.25 (SD: 0.06)-cm silicon substrates. Furthermore, it was confirmed that our method is suitable for ribbons derived from water-soluble resin blocks. We were also able to stain the specimens by post-embedding immunocytochemistry. Thus, our method is useful in mounting numerus sections on a substrate for array tomography with SEM.
Phosphatidylinositol 3,4-bisphosphate [PtdIns(3,4)P2] is a phosphoinositide that plays important roles in signal transduction, endocytosis, and cell migration among others. The intracellular distribution of PtdIns(3,4)P2 has mainly been studied by observing the distribution of GFP-tagged PtdIns(3,4)P2-binding protein domains in live cells and by labeling with anti-PtdIns(3,4)P2 antibody in fixed cell samples, but these methods only offer low spatial resolution results and may have pitfalls. In the present study, we developed an electron microscopic method to observe the PtdIns(3,4)P2 distribution using the SDS-treated freeze-fracture replica labeling method. The recombinant GST-tagged pleckstrin homology (PH) domain of TAPP1 was used as the binding probe, and its binding to PtdIns(3,4)P2 in the freeze-fracture replica was confirmed by using liposomes containing different phosphoinositides and by the lack of labeling by a mutant probe, in which one amino acid in the PH domain was substituted. The method was applied to NIH3T3 cell samples and showed that the increase of PtdIns(3,4)P2 in cells treated with hydrogen peroxide occurs in the cytoplasmic leaflet of the plasma membrane, except in the caveolar membrane. The present method can define the distribution of PtdIns(3,4)P2 at a high spatial resolution and will facilitate our understanding of the physiological function of this less studied phosphoinositide.