Platelet interior imaging technologies

To the editor:

Harry F. G. Heijnen and his colleagues recently reported studies of the platelet interior using electron tomography.¹  The methods used to freeze platelets and prepare them for cryo-electron tomography are of interest. However, most of the information presented is not new. A major finding was that the open canalicular system and dense tubular systems of channels were highly intertwined and formed close associations in specialized membrane regions. This observation is not new. Interaction of the 2 channel systems was first reported in 1972,²  and the name “membrane complexes” assigned to them. Their similarity to the sarcoplasmic reticulum of embryonic muscle and ability of the dense tubular system to bind divalent cations suggested a role of membrane complexes in platelet muscle physiology.³ 

A review chapter, “Platelet Structure,” at page 70 (Figures 3-74, 3-75, and 3-76) in the second edition of Michelson's textbook Platelets⁴  provides a more comprehensive picture of membrane complexes. Identification of α granule subtypes included organelles containing the multimeric von Willebrand factor assemblies in tubules eccentrically located in the α granule matrix.¹  Heijnen and colleagues cite a paper by Cramer et al⁵  for recognizing the tubular elements in α granules, demonstrating that they are von Willebrand factor multimers similar to Weibel-Palade bodies. However, an earlier study in 1968⁶  described them in thin platelet sections by transmission electron microscopy (TEM). The authors of the current paper¹  frequently encountered α granules with an elongated, “tubular shape,” and their Figures 4 and 5 provide images of the tubular subtypes. However, images in Figure 3-49 on page 62 of Michelson's text⁴  provide an image of α granules with extensions taken by TEM that is clearer than provided by tomography. The extensions appear as dense as the granules from which they originate. Therefore, we referred to them as “rod-like,” rather than “tubular.” The presence of cross-striations in the rods also was noted in another study.⁷ 

Electron tomography or its interpretation may make errors in organelle identification. The tomographic slices of a “tubular” α granule in Figure 5C-F and its reconstruction in 5G-H are interesting, but the organelle is not an α granule. It is a δ granule (dense body). Several examples are shown in Michelsen⁴  on page 62 in Figures 3-50 and 3-51, on page 63 in Figure 3-52, and in examples included here (Figure 1). The images in the enclosed illustrations were taken by TEM on whole-mount preparations of normal platelets.

In summary, Heijnen et al have used a useful new technology to review the platelet interior. However, based on the concerns raised in this letter, it does not appear that electron tomography has replaced thin-section and whole-mount TEM.