Forget, for a moment, about the familiar APP mutations that produce excess β amyloid in the brain and cause Alzheimer's disease. There are some mutations that do not cause the full range of AD pathology, despite producing β amyloid peptides. The E693Q mutation, for example, causes a dementing disorder called hereditary cerebral hemorrhage with amyloidosis-Dutch type (HCHWA-D). In these patients, amyloid build-up is predominantly vascular, and a report in this month’s Nature Neuroscience suggests that the difference between brain parenchymal amyloidosis and cerebral amyloid angiopathy is all in the Aβ40/Aβ42 ratio.

In their new transgenic mouse model of HCSWA-D, Mathias Jucker and first author Martin Herzig of the University of Tübingen in Germany, along with collaborators from several other institutions, find that the E693Q mutants recapitulate the human condition by accumulating mostly vascular amyloid. Significantly, Herzig found that the animals produce mainly Aβ40, and when the authors examined autopsy samples from HCHWA-D patients, they found that they, too, had more Aβ40 than Aβ42—18 times more.

The evidence supports the notion that different Aβ species have very different interactions with their extracellular environments, and that parenchymal amyloid build-up requires Aβ42 seeding, while a surfeit of Aβ40 can drive vascular amyloid build-up. This finding leads Herzig and colleagues to raise the concern that therapeutic approaches based on reducing Aβ42 could have the unintended effect of promoting Aβ40-driven vascular amyloidosis. In support of this, Herzig found that crossing the E693Q transgenic mice with a strain harboring a presenilin-1 mutant that preferentially cleaves AβPP at the 42 position, results in offspring that make predominantly Aβ42 and produce amyloid mainly in the brain parenchyma. This suggests that the reverse, promoting Aβ40 over Aβ42, could cause vascular amyloidoses.

Beyond these insights into the effects of favoring different Aβ species, studies in this new animal model should prove very valuable, because vascular amyloidosis may be a contributing factor to cognitive decline both in AD and in normal aging.—Hakon Heimer


  1. The exciting data presented in the Herzig et al. paper provide key insights into some of the mechanistic differences in the deposition of the Aβ peptide in brain parenchyma vs. cerebrovasculature. Hereditary cerebral hemorrhage with amyloidosis-Dutch type is an autosomal dominant form of cerebral amyloid angiopathy (CAA) resulting from a mutation within the Aβ coding region at amino acid 22. Expression of the human APP-Dutch transgene under control of the neuron-specific Thy1.2 promoter in mice results in almost exclusive deposition of the Aβ peptide in the cerebrovasculature, leading to smooth muscle cell degeneration, hemorrhage, and inflammation. Thus, this model recapitulates many key aspects of the human disease. Expression of wild-type human APP transgene resulted in mostly parenchymal plaque deposition as has been seen with expression of many of the APP mutations that result in familial AD. They also show that APP-Dutch mice have a higher ratio of Aβ40:Aβ42 as compared to the wild-type human APP mice. Crossing the APP-Dutch mice to mice expressing a mutant presenilin-1 gene (a mutation known to favor generation of Aβ42) results in a striking redistribution of amyloid pathology from the cerebrovasculature to the parenchyma.

    Previous studies have suggested that a higher ratio of Aβ40:Aβ42 promotes the formation of CAA. To date, the data in this paper provide the best mechanistic evidence in vivo that the ratio of Aβ40:Aβ42 is a key determinant of whether Aβ deposits in brain parenchyma vs. the cerebrovasculature. These data also suggest that therapeutics that decrease the amount of Aβ42 in the brain may result in a reduction of parenchymal amyloid plaques at the expense of generating an increase in the amount of CAA by simply altering the ratio of Aβ40:Aβ42. Further studies of the effects of potential therapeutics in mouse models known to produce significant CAA vs. models known to have much less CAA will be necessary to sort out these possibilities.

  2. The work provides insights into the significance of conformational versus metabolic properties in the occurrence of pathological phenotypes, i.e., amyloid angiopathy versus parenchymal amyloidosis.

    View all comments by Takaomi Saido
  3. The beautiful series of papers by Jucker and colleagues begin to paint a picture of how cerebrovascular amyloid deposition (or CAA) might occur:

    1. Neuronally produced Aβ peptide appears to have the opportunity to deposit in the brain parenchyma as plaques, to reach the vessel wall and perivascular space, and deposit as CAA, or to be cleared without depositing.

    2. Various factors, including amount of Aβ, ratio of Aβ42:Aβ40, and pathogenic mutations within the Aβ sequence (among a host of factors involved in Aβ deposition and clearance) can predispose Aβ to deposit or be cleared.

    3. For extensive CAA to occur, these amyloidogenic factors need to happen in exquisitely fine balance. Too much "amyloidogenicity" causes Aβ to deposit primarily as plaques; too little, and it is cleared without depositing. The E693Q APP transgenic mouse described in this paper appears to hit the CAA "sweet spot" with almost exclusive vascular deposition; addition of mutant presenilin tips the balance squarely towards plaque deposition.

    Though clearly an oversimplification, it will be interesting to see if this model makes useful predictions as further data from transgenic mice emerge—and as they are translated into therapeutic approaches to human AD and CAA.

Make a Comment

To make a comment you must login or register.


No Available References

Further Reading

Primary Papers

  1. . Abeta is targeted to the vasculature in a mouse model of hereditary cerebral hemorrhage with amyloidosis. Nat Neurosci. 2004 Sep;7(9):954-60. PubMed.