Tomograms reveal particles (red) inside the lumen of microtubules (green).
Nearly half a century ago, early microscopic evidence suggested that some sort of particle resided inside microtubules. Many researchers, however, considered them possible artifacts of fixation or staining procedures.
Cryoelectron tomography bypasses most of these problem-inducing steps. With just a quick freeze of intact neuronal cells, tomograms at 5-6–nm resolution revealed globular densities about the size of several tubulin subunits within neurite microtubules.
The globules contact the luminal face of the tubulin subunits, but higher (∼2 nm) resolution will be needed to identify precise interaction sites. The particles were found more often in depolymerizing microtubules and thus might increase microtubule dynamics. So far, densities have only been seen in the microtubules of astrocytes, neurons, and neuronal precursors, which have particularly long and dynamic microtubules.
As the internal face is predicted to harbor domains that can be acetylated, the authors suggest that the particles might be acetyltransferases or deacetylases. Other possibilities remain, however. Perhaps they are proteins or mRNAs that are transported or simply stored within microtubules, to be released upon the filament's depolymerization. To identify these globules, intact neuronal microtubules will first have to be isolated. 
