First author Yvonne Eisele and colleagues induced amyloid deposits in the brains of young APP23 transgenic mice. These animals express mutant human amyloid precursor protein and begin to form amyloid plaques around the age of six months but don’t have widespread pathology until much later. Eisele and colleagues used diluted brain extracts from aged APP23 animals as seeds and injected them into various brain regions of young (two- to five-month-old) mice. Looking at the brains three to six months later the researchers found that the APP23 brain extract, but not extract from wild-type mice, induced β amyloidosis in the hippocampus, parietal cortex, entorhinal cortex, striatum, and the olfactory bulb.

The amyloidosis took time to propagate. At first the brains looked normal, but within three months of injection Aβ deposits were found near the injection site. After another three months the pathology spread to adjacent sites, possibly by propagation along neural projections. Injections in the entorhinal cortex led to Aβ deposits in the molecular layer of the dentate gyrus, for example, while striatal injections induced deposits in connected areas of the neocortex. There were regional differences in the type of deposit formed as well. In the entorhinal cortex and hippocampus deposits were congophilic, suggestive of senile plaques, whereas only diffuse immunoreactive deposits were found in the striatum. This is the typical pattern that develops in these mice naturally, which perhaps indicates that the type of Aβ deposited is independent of the seed that induces it. But Jucker stressed that the situation is more complex. His previous work has shown that the relative amounts of Aβ42 and Aβ40 can influence the type of deposit induced by Aβ extracts (see ARF related news story on Meyer-Luehmann et al., 2006), and he thinks that region-specific differences in the Aβ40/42 ratio could play a role. “I don’t think anybody has really looked at this, but I’d say that is one possibility. The other possibility is just that the region-specific amyloid formation is a result of regional differences in brain structure,” he told ARF. Jucker said that in transgenic mice the explanation for different types of deposit could also hypothetically lie in the strength of transgene expression, but that explanation is even more unlikely in humans because APP expression is similar throughout many brain structures.

To see if Aβ seeds are infectious, the researchers used exactly the same type of protocols that demonstrate infectivity of prion particles in mice. Eisele and colleagues introduced APP23 brain extracts into the periphery. Injecting via oral, intraocular, or intranasal routes failed to induce amyloidosis in the brain up to eight months later. The work indicates that amyloid extracts are not infectious (for a review on this topic, see Walker et al., 2006), though Jucker said there are two caveats to that conclusion. "One is that the incubation times may need to be longer, and the second is that APP transgene expression in APP23 mice is largely limited to the brain, whereas peripheral PrP is a prerequisite for PrP infectivity,” said Jucker. If the same was true for Aβ, then it would be hard to prove infection in these APP23 mice because transgene expression is limited to the brain. Jucker said they now have mice with transgene expression in the periphery and he plans to test those animals in the same manner.

Perhaps of clinical relevance, since metal surgical instruments are often re-used after sterilization, the researchers found that stainless steel wires dipped in brain extract caused amyloidosis when placed into the brain, and the amount of Aβ on the wires was less than needed to induce Aβ deposits when injected in soluble form. One possible explanation for this is that immobilization on the wires extends the life of the Aβ seed, suggest the authors. Boiling the wires did not prevent the induction of amyloidosis. “Thus, transmission of β amyloid-inducing activity to patients through the use of neurosurgical instruments is a theoretical possibility,” write the authors. But they note that such transmission has never been documented, though it may be masked by very long incubation times.

Though the paper supports the overwhelming epidemiological evidence that amyloidoses are not infectious diseases, it does speak to the greater debate about host to graft transmission, which could scupper potentially valuable cell replacement strategies. Fetal neurons transplanted into Parkinson patients to boost dopaminergic output were found to contain Lewy bodies years later, for example, raising the specter of toxic α-synuclein transmission (see ARF related news story). A similar strategy was tried in Huntington disease patients, and in this week’s PNAS, researchers report that grafted cells also underwent neurodegeneration (see Cicchetti et al., and related Q&A with authors on PD Online Research). In this case the affected neurons showed no signs of abnormal huntingtin deposits, suggesting that the host environment rather than transmission of huntingtin toxicity might be to blame. In some cases the spoiler may be the barrel, not the apples.—Tom Fagan.

Reference:
Eisele YS, Bolmont T, Heikenwalder M, Langer F, Jacobson L, Yan Z-X, Roth K, Aguzzi A, Staufenbiel M, Walker LC, Jucker M. Induction of cerebral beta-amyloidosis: Intracerebral versus systemic Abeta inoculation. PNAS online 2009, July 21. Abstract