Updated 1/30/2007 — Note by Alzforum Editor: The Doglio et al. paper has been retracted at the request of the authors. According to the authors, several of the figures were found to contain serious inaccuracies and no longer support the major conclusions of the paper. Click on retraction page links for complete author statement.

In the middle of a lively debate about how exactly presenilin (PS) mutations cause Alzheimer disease (see our ongoing ARF Forum Discussion) comes a paper that opens up a whole new angle on the issue. Writing in Neuron this week, researchers from the labs of Carlos Dotti at the Catholic University of Leuven, Belgium, Fabian Feiguin at the University of Turin, Italy, and Mark Fortini from the National Cancer Institute, Frederick, Maryland, and their collaborators propose a new presenilin-dependent neurodegenerative mechanism that is totally independent of γ-secretase activity.

The investigators found that presenilins and two other proteins of the γ-secretase complex, nicastrin and Aph-1, are needed to keep the lid on tau hyperphosphorylation in a Drosophila eye model of tauopathy. It seems that presenilin and nicastrin are required for the correct subcellular localization of the tau kinase GSK3β, while Aph-1 regulates the distribution of a different tau kinase, PAR-1. Lowering these γ-secretase components, or introducing a specific, familial Alzheimer disease (FAD) PS mutation, results in kinase mislocalization, high levels of tau phosphorylation, and neurodegeneration. Protease-deficient presenilins can rescue the phenotype, indicating that γ-secretase is not part of the pathology.

On the other side of the coin, a paper in yesterday’s Nature Medicine online reveals a presenilin-dependent neurodegenerative pathway that does require γ-secretase, but not in its typical role of cleaving amyloid precursor protein. That work, from Mark Mattson and colleagues at the National Institute on Aging in Baltimore, Maryland, shows that cleavage of the PS substrate Notch after ischemic brain injury contributes to neuron death. In this case, short-term treatment with γ-secretase inhibitors reduces brain damage and improves functional outcome in a mouse model of focal ischemia.

In the tau study, first author Laura Doglio and colleagues followed up on recent observations suggesting that presenilins directly influence tau hyperphosphorylation. The amyloid cascade hypothesis places tau hyperphosphorylation and neurofibrillary tangle formation downstream of amyloid peptide production, but several observations have called that ordering into question. In vitro studies show that PS suppresses tau phosphorylation by regulating the PI3 kinase/AKT/GSK3β cascade (Baki et al., 2004). This regulation doesn’t require γ-secretase activity and is disrupted by FAD mutations, which result in tau hyperphosphorylation. These results could explain the surprising phenotype of PS conditional knockout mice, which have no amyloid, but do have hyperphosphorylated tau and neurodegeneration (see ARF related news story). In humans, PS mutations have been identified that cause frontotemporal dementia with tauopathy in the absence of amyloid accumulation (Dermaut et al., 2004).

To trace the route from presenilins to tauopathy in vivo, the researchers took advantage of a Drosophila model of tauopathy created by coauthor George Jackson. In their model, eye-specific expression of human tau causes some disorganization and degeneration of retinal photoreceptor neurons (see ARF related news story). When they replaced one copy of PS with a null allele, the tau phenotype was accentuated. At the same time, tau phosphorylation increased and solubility decreased.

Further genetic manipulations and biochemical measures implicated shaggy, the Drosophila equivalent of GSK3β, as the kinase responsible for some of the tau hyperphosphorylation and neurodegeneration. Overexpression of shaggy caused mild retinal degeneration, which was enhanced by PS1 reduction, while removal of endogenous shaggy suppressed the PS1 phenotype. Western blots confirmed that shaggy was activated by phosphorylation in PS-deficient flies. These results suggest that PS reins in tau phosphorylation by keeping shaggy/GSK3β turned off. Consistent with this idea, removing the upstream inhibitory regulators of shaggy in the PI3K/Akt pathway increased tau phosphorylation and neurodegeneration, while activating them suppressed the effects of PS deficiency.

A clue to how PS regulates shaggy came from studies on the subcellular localization of the enzyme—in the normal eye, shaggy was localized in discrete vesicles close to cell junctions, while in PS-deficient cells, the kinase adopted a more diffuse pattern.

All of these effects were reversible by adding back wild-type PS, but not a FAD mutant allele with an exon 9 deletion. Surprisingly, two PS alleles lacking γ-secretase catalytic activity (Asp-257 or Asp-385 mutations) performed as well as wild-type in the add-back experiments. Consistent with this result, the investigators showed that γ-secretase inhibitors had no effect on tau phosphorylation or the activation status of shaggy.

Other components of the γ-secretase complex also affected tau phosphorylation. Either nicastrin or Aph-1 mutants enhanced the tau neurodegeneration phenotypes, with both boosting the phosphorylation and insolubility of phospho-tau. Further investigation showed that nicastrin, like PS, was required for proper compartmentalization of shaggy/GSK3β kinase. Aph-1, on the other hand, regulated the localization and activation of a PKC, the activator of the tau kinase PAR-1.

How might PS regulate kinase localization? Possibly, direct association of the proteins could be involved, since the PS mutation that affected tau phosphorylation was missing a loop thought to be important for interactions with other proteins. It will be critical to test additional mutations to see if exacerbated tau phosphorylation is a result of all, some, or just a few of them.

One strength of this study lies in the fact that the flies develop neurodegeneration with a partial loss of presenilin, with only one allele knocked out, a situation comparable to what happens in humans with PS mutations. If a normal role of PS is to regulate tau kinases to keep phosphorylation in check, then this fits with the idea that FAD PS mutations cause a partial loss of function, which shows up even in the presence of one normal allele.

With more study, presenilins keep on revealing new functions, complicating efforts to develop therapeutics targeting their actions. In the case of γ-secretase inhibitors, the effects of long-term inhibition of processing of substrates other than APP are a great concern. But for one substrate, Notch, short-term inhibition might turn out to be useful in treating stroke.

That’s the message from a study that looked at the effects of inhibiting γ-secretase on outcome of ischemic brain injury in mice. A trio of first authors, Thiruma Arumugam, Sic Chan, and Dong-Gyu Jo joined forces to show that focal ischemia induced both γ-secretase activity and Notch cleavage in mouse brain. Treating mice with the γ-secretase inhibitor DAPT shut down Notch cleavage and reduced the area of damaged tissue after infarct. Using cell assays and Notch-deficient mice, they showed that production of the Notch intracellular domain (NICD) fragment resulted in activation of apoptotic pathways in neurons, activation of microglia, and infiltration of lymphocytes into the damaged tissues. All these activities were reduced, and neurological scores increased, when the γ-secretase inhibitors were given shortly before the blood supply was cut off, or up to four hours after reperfusion. The fact that targeting NICD production affects not just neurons but multiple cells types in affected tissues suggests that γ-secretase inhibitors could be superior to therapeutics that only affect neurons, the authors conclude.—Pat McCaffrey


  1. The paper by Doglio et al. has important implications for both the biological functions of the presenilin (PS) complex and the mechanisms by which PS FAD mutations promote neurodegeneration and other structural abnormalities in AD. The paper verifies previous results that presenilins regulate the PI3K/Akt cell survival pathway through a mechanism that is independent of the γ-secretase activity (Baki et al., 2004). This report is also consistent with recent evidence that FAD mutations may promote cell death and AD-related tau overphosphorylation through a mechanism that involves loss of presenilin function in the PI3K/Akt/GSK3 and other cell signaling pathways (Baki et al., 2004; Kang et al., 2005). This last concept is in contrast to the widely held view that FAD mutations cause a gain of presenilin γ-secretase function thus increasing production of neurotoxic species like Aβ1-42 (see, however, ARF ongoing discussion). The data on Aph-1/aPKC/PAR-1 indicate a new pathway that may also contribute to tau phosphorylation. Clearly, new exciting data emerging in the last two years from several laboratories promote novel concepts not only for the biological functions of PS1, but also for the mechanisms by which presenilin FAD mutations induce the specific neuropathology associated with AD.


    . gamma-cleavage-independent functions of presenilin, nicastrin, and Aph-1 regulate cell-junction organization and prevent tau toxicity in vivo. Neuron. 2006 May 4;50(3):359-75. PubMed. RETRACTED

    . PS1 activates PI3K thus inhibiting GSK-3 activity and tau overphosphorylation: effects of FAD mutations. EMBO J. 2004 Jul 7;23(13):2586-96. PubMed.

    . Presenilins mediate phosphatidylinositol 3-kinase/AKT and ERK activation via select signaling receptors. Selectivity of PS2 in platelet-derived growth factor signaling. J Biol Chem. 2005 Sep 9;280(36):31537-47. PubMed.

  2. This is a very interesting paper, which further extends our understanding about PS1-induced toxic gain of function(s). Indeed, increased GSK3 activation (through serine 9 dephosphorylation) has previously been reported in PS1 animal models (Pigino et al.) and cultured neuronal cells (Takashima et al., Irving et al.). However, these studies did not address the role of other γ-secretase proteins such as Nicastrin, and Aph-1, an issue nicely addressed by the studies in Doglio et al. What remains controversial from these and similar studies is the role of tau protein within this mechanism. Although tau is clearly affected as a downstream component in the PS1-GSK3 pathway, it is possible that tau represents one of many more GSK3 targets. This and other studies need to consider the possibility that deregulation of well-characterized GSK3 substrates other than tau (i.e., kinesin-1, CRMP-2, etc.) might be critical components leading to selective neuronal toxicity.


    . Alzheimer's presenilin 1 mutations impair kinesin-based axonal transport. J Neurosci. 2003 Jun 1;23(11):4499-508. PubMed.

    . Tau phosphorylation in cells transfected with wild-type or an Alzheimer's disease mutant Presenilin 1. Neurosci Lett. 1997 Jan 31;222(2):71-4. PubMed.

    . Presenilin 1 associates with glycogen synthase kinase-3beta and its substrate tau. Proc Natl Acad Sci U S A. 1998 Aug 4;95(16):9637-41. PubMed.

    View all comments by Gerardo Morfini
  3. Note by Alzforum Editor: This paper has been retracted. According to the authors, several of the figures were found to contain serious inaccuracies and no longer support the major conclusions of the paper. Click on retraction page links for complete author statement.

    View all comments by Gabrielle Strobel
  4. I am saddened and disappointed that irregularities in the figures of this manuscript led to the retraction of this paper. Accordingly, I formally retract my prior commentary discussing this study. We are all reminded to apply the greatest care and integrity to our research.

    View all comments by Gerardo Morfini

Make a Comment

To make a comment you must login or register.


Webinar Citations

  1. Gain or Loss of Function—Time to Shake up Assumptions on γ-Secretase in Alzheimer Disease?

News Citations

  1. The Senility-Presenilin Connection Turned Upside Down
  2. Wingless Pathway Helps Tauopathy Take Off

Paper Citations

  1. . PS1 activates PI3K thus inhibiting GSK-3 activity and tau overphosphorylation: effects of FAD mutations. EMBO J. 2004 Jul 7;23(13):2586-96. PubMed.
  2. . A novel presenilin 1 mutation associated with Pick's disease but not beta-amyloid plaques. Ann Neurol. 2004 May;55(5):617-26. PubMed.

External Citations

  1. retracted

Further Reading

No Available Further Reading

Primary Papers

  1. . gamma-cleavage-independent functions of presenilin, nicastrin, and Aph-1 regulate cell-junction organization and prevent tau toxicity in vivo. Neuron. 2006 May 4;50(3):359-75. PubMed. RETRACTED
  2. . Gamma secretase-mediated Notch signaling worsens brain damage and functional outcome in ischemic stroke. Nat Med. 2006 Jun;12(6):621-3. PubMed.