When molecular biologists go “fishing,” they don’t really expect to catch any fish. Not unless you’re talking about the protein FISH, named after its five SH3 domains. In the February 22 PNAS, Irene Griswold-Prenner and colleagues from Elan Pharmaceuticals, San Francisco, report catching this particular one while trying to net proteins that might mediate Aβ neurotoxicity. The paper is available in this week’s early online edition of the journal.

Griswold-Prenner’s fishing expedition started, like so many others, with some bait, in this case commercially available antibodies to phosphotyrosine antigens. Back in 2001, researchers from Wyeth-Ayerst (see Kajowski et al., 2001) had found that Aβ binds to a G protein coupled receptor coined β amyloid binding protein (BBP). G proteins, of course, are known to activate tyrosine kinases, so the Elan group wondered if Aβ was triggering protein phosphorylation. To test this, they used the antibodies to identify potential downstream targets of BBP or other Aβ-activated kinases.

First author Nikolay Malinin and colleagues used the antibodies to test human cortical cultures—arguably a more pathologically relevant cell-based model for Aβ toxicity than cell lines or non-human tissues. They found that one antibody—raised to a phosphorylated form of the epidermal growth factor receptor (EGFR) —detected increased phosphorylation of a single protein when cortical cultures were treated with a toxic form of Aβ. But the protein, with a molecular weight of 165 kDa, was too big to be the EGFR, and when Malinin and colleagues identified it, they found it was, in fact, the FISH adapter protein.

How might FISH play a role in Aβ-mediated toxicity? The protein is normally kept in an inactive state, brought about by intramolecular binding between an N-terminal PX domain and one of its SH3 domains. Malinin and colleagues found that expression of FISH minus the PX motif led to neurotoxicity even in the absence of Aβ. By contrast, overexpressing the PX-containing N-terminal end of the protein had the opposite effect, protecting the cortical cultures from toxic Aβ.

FISH is known to interact with the ADAMs family of disintegrins/metalloproteinases, sometimes known as “sheddases” because they cleave extracellular domains from a number of membrane proteins. The ADAMs are a self-destructive bunch, cleaving themselves in addition to their other targets. FISH is known to interact with ADAM12, 15, and 19, and this prompted the authors to examine the role of the ADAM proteins in Aβ-mediated toxicity. When Malinin exposed human cortical cultures to toxic forms of Aβ, he found that only levels of ADAM12 were reduced, suggesting that at least this metalloproteinase had been activated by FISH. Further experiments supported this idea. Deletion of the ADAM-binding domain in FISH prevented Aβ toxicity, as did expression of a metalloproteinase-deficient ADAM12.

Many other proteins and signaling pathways have been implicated in mediating Aβ toxicity; studies include work in cultured rat neurons (see ARF related news story) and in human neuroblastoma cells (see Wang et al., 2000), to quote but two examples. No consensus has emerged, however, on which players and pathways are the key ones. Where, then, do FISH and ADAM12 fit in? And how would they relate, if at all, to ADAM10, considered to be the α-secretase that generates a protective snippet of APP (see ARF related news story)?

Finding a solid in-vivo correlation between Alzheimer disease and the FISH/ADAM pathway would go a good way toward establishing their importance. As a first step toward that goal, the authors report that ADAM12 appears to be reduced in tissue from AD brain, perhaps in a sign of autocatalytic degradation, but the reduction is modest, the sample size small, and autopsy studies come with their own limitations. Nevertheless, the authors write that their findings “show an unusual pathway of Aβ-induced neurotoxicity through FISH-mediated regulation of ADAM(s) and suggest a previously undescribed strategy for AD treatment by targeting the metalloprotease activity of ADAM(s).”—Tom Fagan


  1. This paper provides insight into a potentially novel mechanism of Aβ toxicity. The authors start by looking for proteins in which phosphorylation is induced by toxic levels of Aβ in human cortical cultures. They identify the FISH (for Five SH3 domains) protein as being phosphorylated within one hour of Aβ treatment, indicating that this is an early event induced by Aβ. FISH can bind to the ADAM family of disintegrin metalloproteases via one of its SH3 domains, leading to activation of the ADAMs protease activity. Phosphorylation of FISH leads to a change in its distribution from the cytoplasmic to perinuclear region, allowing it to activate ADAM. Expression of a FISH construct without the ADAM binding SH3 domain protects against Aβ toxicity, suggesting FISH activation of ADAM is important in this process. The authors then focus on ADAM12, a known target of FISH, and show that ADAM12 may be activated by Aβ expression of an ADAM12 construct that lacks the protease domain protects against Aβ induced cell death, suggesting that ADAM12s protease activity has an important role.

    Two fundamental questions remain: How does Aβ lead to FISH phosphorylation and how does increased ADAM activity lead to cell death? The ADAM family is implicated in the processing of many transmembrane proteins including cytokines, growth factors, and of course APP. Does ADAM12 cleave a specific substrate that leads to cell death? It is not clear if ADAM12 can cleave APP at the α site (ADAMs 9, 10, and 17 are implicated so far). ADAM17 can release TNFα, which can lead to cell death via activation of the TNFα receptor signaling pathway. It would be interesting to determine if ADAM12 also has this activity, as TNFα is increased in Alzheimer disease brain and the signaling pathway has been implicated in Aβ toxicity (Li et al., 2004).

    Aβ toxicity can also be modulated by factors that protect against a range of cellular insults, for example, caspase inhibitors and antioxidants. It is not clear if the pathway described here is related to these global protective pathways or is selective for Aβ toxicity. It would thus be useful to know if ADAM12 modulates free radical toxicity, serum-deprivation-induced apoptosis, glutamate toxicity, etc.


    . Tumor necrosis factor death receptor signaling cascade is required for amyloid-beta protein-induced neuron death. J Neurosci. 2004 Feb 18;24(7):1760-71. PubMed.

  2. I think we are getting off track abit. It has been established for some time that various phosphorylation processes occur in the presence of amyloid beta, including tau protein, tyrosine kinase receptors, involvement of g protins, etc... amongst others. These observations made here in this paper are interesting, but are not going to help in finding a cure for AD.

    Fundamentals of Biochemistry, Molecular Biology of the Cell, Genes7 and Alzforum.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Do Aβ and APP Team up to Kill Neurons?
  2. α-Secretase Returns to Center Stage

Paper Citations

  1. . beta -Amyloid peptide-induced apoptosis regulated by a novel protein containing a g protein activation module. J Biol Chem. 2001 Jun 1;276(22):18748-56. PubMed.
  2. . beta-Amyloid(1-42) binds to alpha7 nicotinic acetylcholine receptor with high affinity. Implications for Alzheimer's disease pathology. J Biol Chem. 2000 Feb 25;275(8):5626-32. PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. . Amyloid-beta neurotoxicity is mediated by FISH adapter protein and ADAM12 metalloprotease activity. Proc Natl Acad Sci U S A. 2005 Feb 22;102(8):3058-63. PubMed.