As the roster of substrates for the γ-secretase protease continues to swell, the presenilin (PS) proteins seem to be popping up everywhere. Now, Hui Zheng and colleagues from Baylor College of Medicine in Houston, Texas, report they’ve spotted PS in yet another, entirely new pathway. Their work, published December 29 in PNAS Early Edition, shows that γ-secretase is required for melanin synthesis in the eyes and skin, at least in mice. Their results show that in the absence of γ-secretase activity, the intracellular trafficking of the melanin-producing enzyme tyrosinase to melanosomes stops, preventing pigment production. Tyrosinase and several related melanosomal type I membrane proteins appear to be substrates for γ-secretase, and in mice, the presenilin 1 M157V familial Alzheimer disease mutation impairs melanin synthesis. With this study, Zheng et al. have uncovered several new substrates for the γ-secretase, and a new role for PS in intracellular protein transport.

The initial clue linking presenilins to melanin production has been staring researchers in the face for several years. In previous work, Zheng and colleagues attempted to rescue the embryonic lethal phenotype of PS1/PS2 double knockout mice by breeding in a human PS1 transgene whose expression was restricted to the central nervous system. The approach was partially successful, and rather than perishing as embryos, the animals at least survived to birth. They expired soon after, however, and displayed massively abnormal kidney development due to compromised Notch signaling in non-neural tissues (Wang et al., 2003).

In those experiments, the researchers found they could distinguish the rescued mice from their control littermates at a glance—the transgene recipients all had white eyes. In the new work, chasing the white eye phenotype led first author Runsheng Wang and colleagues directly to the melanin synthesis pathway. This black pigment is produced inside melanosomes, membrane-bound organelles that contain the enzyme tyrosinase, which catalyzes the rate-limiting step in the transformation of tyrosine into melanin. [If this sounds familiar, it is. Melanosomes recently grabbed the attention of AD researchers, who found them to be filled with an amyloid of the Pmel17 protein, the first physiological amyloid found in higher organisms (see ARF related news story)].

With no PS1 expression, the eyes of the rescue mice were morphologically normal, but lacked pigment granules in the retinal pigmented epithelium (RPE) cells. The researchers showed that the mice had normal levels of the tyrosinase and two associated proteins, but the tyrosinase was mislocalized. PS+ cells contained a small number of tyrosinase-containing vesicles located near the trans-Golgi network. In normal eye cells, these vesicles appear to shuttle tyrosinase from where it is made in the secretory pathway to nearby early-stage melanosomes. But in the PS-null cells, these small, tyrosinase-loaded vesicles accumulated, and could be seen all over the cell body. The same cells contained only early-stage melanosomes that failed to mature or produce melanin.

Successful melanin synthesis depends on PS in skin cells, as well. Blocking γ-secretase activity with the inhibitor DAPT caused skin melanocytes to behave like the PS-null cells: Melanin synthesis was blocked and melanosome maturation stalled.

For many γ-secretase substrates, inactivation of PS leads to accumulation of a C-terminal cleavage fragment. Tyrosinase, as a type I membrane protein, was a candidate substrate, and the researchers detected a 7 kDa C-terminal fragment in either in PS 1 -/- cells, or in cells treated with DAPT. A CTF was also detected for two other tyrosinase-associated, melanosome-targeted proteins after DAPT treatment, suggesting that all three proteins could be PS substrates.

But what about the effect of FAD mutations on melanin production? From previous work (Wang et al., 2004), Zheng and colleagues knew that animals carrying the M146V FAD allele displayed learning and neurogenesis phenotypes, but only when the wild-type PS1 allele was absent. To look at effects on pigment in vivo, they bred heterozygous PS1 knockouts with knock-ins bearing the M146V FAD mutation in the endogenous PS1 gene, all on a PS2 null background. The mice carrying only the FAD mutant gene (PS2-/-, PS1 M146V/-), had a distinctly lighter coat color than did mice with a wild-type PS1 gene (PS2-/-, PS1+/-). PS1 gene dosage also appeared to affect pigmentation, although to a lesser extent than the mutation. Thus, while there was no noticeable difference in coat color between wild-type mice (PS1+/+) and PS1 knockout heterozygotes (PS1+/-), visual inspection of tail skin pigmentation and biochemical measurements of melanin content revealed that the PS1+/- produced less melanin than did the PS1 wild-type, and the PS1 M146V/- produced the least.

The results of Wang et al. establish a new role for PS in protein trafficking to melanosomes, but the mechanistic details of this regulation remain to be worked out. The physiological significance of the cleavage of tyrosinase cleavage and related proteins by γ-secretase is likewise totally unknown. The systems developed by the investigators provide a good opportunity for future studies to untangle these issues and the role of FAD mutations in presenilin function.

With this new data, melanosome biosynthesis is now known to involve not only a novel amyloid, but also a presenilin. A coincidence? one might ask. Or is there a link between PS and the Pmel17 protein, the precursor to melanosome amyloid? According to Michael Marks of the University of Pennsylvania in Philadelphia, it is unlikely that PS plays a role in the processing of Pmel17. Zheng’s data showed that Pmel17 protein levels and immunostaining patterns were unaffected by PS deficiency. In addition, the PS-deficient cells generate relatively intact early-stage melanosomes, which also argues against an obligatory role of PS in Pmel17 amyloid generation, Marks notes.

In the end, getting out of melanosomes and back to AD, Wang et al. conclude, “These findings raise the intriguing possibility that a compromised post-Golgi vesicle transport may contribute to Alzheimer’s disease pathogenesis.”—Pat McCaffrey


  1. The lack of pigmentation in the presenilin mutants is a very interesting and potentially very important finding. It raises the question of what the molecular basis for the dysfunctional melanosomes could be. While the appearance of tyrosinase cleavage fragments upon presenilin inhibition is a useful molecular marker, it is unclear how this relates to the trafficking defect observed; indeed, it would seem more likely to be a consequence of tyrosinase mistrafficking rather than a cause. Cleavage of a vesicle "cargo" protein would be unlikely to cause the accumulation of vesicles. Moreover, it is unclear how loss of presenilin proteolytic activity would lead to increased proteolytic cleavage of tyrosinase. Thus, the mechanistic link between presenilin activity and melanosome targeting remains to be solved.

    While cleavage of the pigment cell-specific protein Pmel17 by a proprotein convertase is necessary for amyloidogenesis and formation of the fibrous melanosome matrix, to my knowledge there is no link between presenilins and Pmel17 trafficking or processing. Moreover, we know from unpublished work on the mouse silver mutant that a deficiency of Pmel17 has no effect on the segregation of melanosomes from endosomes/lysosomes or on the targeting of late melanosome cargo like tyrosinase, Tyrp1, or DCT, as Wang et al. see in their PS1-deficient cells. The one potential link between Pmel17 and presenilins may be the fate of the small, membrane-bound Mβ cleavage fragment after the larger lumenal Mα fragment undergoes amyloidogenesis. We know that we can partially stabilize Mβ in melanocytic cells by treatment with lysosomal protease inhibitors (unpublished data), but the effect is not dramatic. Thus, it is quite possible that presenilins have something to do with the elimination of this fragment. However, images shown in Wang et al. suggest that while such an activity of presenilins may facilitate optimal Mα amyloidogenesis, it is not likely to be required. The appearance of relatively intact stage II melanosomes would argue against a requisite role for PS1/2 activity in the Pmel17 life cycle.

    It is perhaps surprising that the authors did not note the morphological similarity of PS1-deficient cells with another pigmentation mutant, the p or OCA2 mutation. P encodes a 12-transmembrane domain protein with similarity to bacterial anion permeases, and some evidence suggests that it localizes to melanosomes. P-deficient (OCA2) melanocytes show defective maturation and trafficking of tyrosinase and Tyrp1, nearly complete lack of pigmentation, and slightly smaller stage II melanosomes as observed upon inhibition of PS1 (for review, see Brilliant, 2001). They also show abnormalities in compartmental pH regulation, which may have a more direct effect on the specialization of the endocytic pathway needed to maintain cargo segregation for melanosome biogenesis (Raposo et al., 2001). Indeed, I find the similarity of PS1-deficient cells with OCA2 cells much more striking than that with the HPS3 cells noted by the authors. P is one of a growing list of ion permeases/transporters that are gaining increasing notoriety for their role in pigmentation, including the underwhite/OCA4 gene (Newton et al., 2001) and more recently SLC24A5 (Lamason et al., 2005). If I were to guess, it may be regulation of one of these critical proteins that is the major target of PS1/2 activity in melanocytes. Of course, it is also highly possible that the target is a SNARE protein or some other regulator of membrane fusion (Chen and Scheller, 2001), but it would be harder to imagine how a defect in such a target would not also cause problems with many other cell types in the conditional knockout.


    . The mouse p (pink-eyed dilution) and human P genes, oculocutaneous albinism type 2 (OCA2), and melanosomal pH. Pigment Cell Res. 2001 Apr;14(2):86-93. PubMed.

    . Distinct protein sorting and localization to premelanosomes, melanosomes, and lysosomes in pigmented melanocytic cells. J Cell Biol. 2001 Feb 19;152(4):809-24. PubMed.

    . Mutations in the human orthologue of the mouse underwhite gene (uw) underlie a new form of oculocutaneous albinism, OCA4. Am J Hum Genet. 2001 Nov;69(5):981-8. PubMed.

    . SLC24A5, a putative cation exchanger, affects pigmentation in zebrafish and humans. Science. 2005 Dec 16;310(5755):1782-6. PubMed.

    . SNARE-mediated membrane fusion. Nat Rev Mol Cell Biol. 2001 Feb;2(2):98-106. PubMed.

  2. Remarkable advances in understanding the pathobiology and mechanisms of neurodegeneration in AD over the past 20 years have engendered realistic hopes for the discovery of more meaningful and effective new drugs to treat AD. Based on these advances, most targets of drug discovery for AD focus on gains of toxic functions by amyloid deposits in AD such as those formed by fibrillar species of tau and Aβ. There is intense interest in Aβ-focused drug discovery efforts; most attempt to block or reverse Aβ plaque formation, or eliminate excess accumulations of Aβ oligomers and fibrils or aggregates of these pathological forms of Aβ.

    Nonetheless, the onset and progression of AD also could result from loss-of-function abnormalities affecting tau or Aβ. When these proteins misfold, oligomerize, and fibrilize to form amyloid filaments and aggregate, the theory goes, they are sequestered and unable to perform their normal functions. However, a detailed understanding of the normal functions of Aβ, its precursor protein, as well as the presenilins required for γ-secretase cleavage of Aβ is still not available, and this limits drug discovery efforts that target losses of functions in these proteins that might contribute to disease pathogenesis.

    Thus, the importance of this paper lies in bringing into sharper focus the functions of the presenilins by implicating them in tyrosinase processing and trafficking. While the repercussions of impairments in these presenilin functions in the brain, and especially in neurons, will need to be examined further, the study opens up new avenues of research into mechanisms of AD and new possibilities for AD drug discovery.

  3. I believe this reductionist finding will have important implications for future molecular research into melanin transport. This could thus lead ultimately to new clinical treatment modalities for deficient genetic pigment manufacture. However, this will not lead to a new way to fight the battle against A.D. We are nowhere near the ability to do genetic reprogramming on such a grand scale. The ramifications of genetic tinkering in the way that would be necessary to apply these research findings is more than 15 years away in my opinion.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Is It Good for You? Amyloid Shows New Side in Mammalian Cells

Paper Citations

  1. . Presenilins are required for the formation of comma- and S-shaped bodies during nephrogenesis. Development. 2003 Oct;130(20):5019-29. PubMed.
  2. . Presenilin 1 familial Alzheimer's disease mutation leads to defective associative learning and impaired adult neurogenesis. Neuroscience. 2004;126(2):305-12. PubMed.

Further Reading


  1. . Partial loss of presenilins causes seborrheic keratosis and autoimmune disease in mice. Hum Mol Genet. 2004 Jul 1;13(13):1321-31. PubMed.
  2. . Presenilin attenuates receptor-mediated signaling and synaptic function. J Neurosci. 2005 Feb 9;25(6):1540-9. PubMed.
  3. . Presenilin 1: more than just gamma-secretase. Biochem Soc Trans. 2005 Aug;33(Pt 4):559-62. PubMed.
  4. . Presenilin 1 mediates the turnover of telencephalin in hippocampal neurons via an autophagic degradative pathway. J Cell Biol. 2004 Sep 27;166(7):1041-54. PubMed.
  5. . Degradative organelles containing mislocalized alpha-and beta-synuclein proliferate in presenilin-1 null neurons. J Cell Biol. 2004 May 10;165(3):335-46. PubMed.

Primary Papers

  1. . Regulation of tyrosinase trafficking and processing by presenilins: partial loss of function by familial Alzheimer's disease mutation. Proc Natl Acad Sci U S A. 2006 Jan 10;103(2):353-8. PubMed.