That functional, non-pathological amyloids exist has been known for some time. In 2005, researchers led by Jeffery Kelly at Scripps Research Institute in San Diego, California, reported that amyloid fibers of the integral membrane protein, Pmel17, accumulate in melanosomes (see ARF related news story). The fibers bind Congo red and thioflavin S, two amyloidophilic dyes commonly used to detect Aβ deposits in brain tissue. But how and where these melanosome fibrils form was unclear. Now, using high pressure freezing (HPF) to rapidly preserve tissue and a form of microscopy called double tilt 3D electron tomography, Raposo and colleagues have peered into melanocytes to show that Pmel17 amyloid-like fibrils begin to form in MVBs and are fully formed into β-sheet like arrays in the melanosomes.

First author Ilse Hurabin and colleagues characterized melanogenesis in a pigmented melanoma cell line called MNT-1. These cells produce melanosomes in all four described stages of maturation. With HPF and the 3D tomography, Hurabin was able to show that the Pmel17 amyloid fibrils form in non-pigmented stage II melanosomes. The fibrils, which appear as separate tubules in two-dimensional snapshots, are actually revealed as being continuous in one plane by 3D tomography, indicating that they assemble into sheets, just like other amyloids (see movie below). The researchers found that the earliest signs of fibrillogenesis occur in MVBs containing one to four internal vesicles. The amyloid fibrils are closely associated with these vesicles, particularly the larger ones, and as the fibrils grow, the vesicles shrink. This concomitant dynamic appears to be related to maturation of melanosomes.

Amyloid fibrils in multivesicular bodies
This 3D reconstruction shows that amyloid fibrils of the Pmel17 protein form in multivesicular bodies in MNT-1 melanoma cells. 3D tilt tomography shows that the fibrils form sheets and are closely associated with intraluminal vesicles. More movies are available as supplementary information on the PNAS website. [Movie copyright: National Academy of Sciences]

It is not clear if amyloid formation in melanosomes relates to Aβ fibrillogenesis, but the authors suggest that the processes may be similar. They describe some interesting parallels. Pmel17, for example, is first cleaved by a proteolytic pro-hormone convertase to yield a smaller Mα fragment, which appears to be the fibrillogenic form. Cleavage of the type II membrane protein BRI2 by furin, a member of the pro-hormone convertase family, is a necessary step in generating the ABri and ADan amyloids that cause familial British and Danish dementias, respectively (see ARF related news story). MVBs may also be a common link. Aβ reportedly accumulates in these organelles (see Takahashi et al., 2002), and some of them mature into exosomes, which Raposo and colleagues have linked to release of pathogenic prions from the cell (see Fevrier et al., 2004). Kai Simons and colleagues at Max Planck Institute in Dresden, Germany, have reported a similar export route for Aβ (see ARF related news story). “My feeling is that these MVBs are providing a favorable environment for amyloid fibril formation,” said Raposo. “They are slightly acidic and their intra-luminal vesicles are lipid raft-like.” Lipid rafts have been linked to amyloidogenesis via both γ- and β-secretase activity.

“We have many parallel research directions in AD that relate to this work, including the role of autophagy, SorLA, lipid/ApoE biology, synaptic dysfunction, protein aggregation, among others,” said Gouras. Next time you are floating on a raft catching some rays, spare a thought for melanogenesis and how it might offer clues to Aβ fibrillogenesis.—Tom Fagan.

Reference:
Hurbain I, Geerts WJ, Boudier T, Marco S, Verkleij AJ, Marks MS, Raposo G. Electron tomography of early melanosomes: implications for melanogenesis and the generation of fibrillar amyloid sheets. Proc Natl Acad Sci U S A. 2008 Dec 16;105(50):19726-31. Abstract