Conference Index | Round Tables | Selected Abstracts | News Summaries

Mechanisms of Neurodegerative Conditions III

by Brian Cummings

July 20, 1998 A-beta's fibril-formation processes were revealed in a tour de force of atomic force microscopy by Harper (abstract 909). The initial species to be detected are 4 nm globular assemblies which range in size from 1.4 to 14nm. He indicated that these globules are similar in size to that reported for "ADDLs". Following the appearence of the globular assemblies, they see protofibrils which do not exceed 200-300 nm in length. Finally they see the traditional 8-10 nm in diameter and up to 1 micron in length Abeta fibrils. This transformation is very rapid and globular and protofibrils rapidly disappear from the samples. Using Abeta1-40, it takes 4 days for the globules to first appear. Fibrils are first observed around 4 weeks and when they first appear, they are already very long. Comparing the rate of assembly between Abeta40 vs. Abeta42, the Abeta42, as others have already indicated, fibril formation is accelerated. If they divide a sample and seed half with 1% fibrils, they see very rapid fibril growth but no fibers in the other half even after 7 days. Using protofibrils as seeds does not produce as rapid a growth into fibrils. Harper also presented data that low temperature inhibits the formation of fibrils and keeps the peptide in the globular (ADDL-like) state. How small can these ATM studies go? The current tip size is 10nm, but using new single wall carbon tips of 1 nm (available at K-Mart in aisle 4), Harper indicated that we are close to imaging in the 2-5 nm range in biological samples. Not bad.

The take-home conclusions from Walsh's presentation (abstract 910) on the process of fibril formation:. LMW Abeta shows little order by CD, but upon incubation LMW Abeta can aggregate. Protofibrils can dissociate, but this pathway is minor. Protofibrils have electrophysiological effects on cells (although Walsh is careful to point out that this doesn't mean that this is physiological). Protofibrils are toxic to cells, but not as toxic as full fibrils.

Greg Cole gave a nice presentation (abstract 911) on the possible presence of "synapoptosis" in both the AD brain and the Hsiao mouse. He has generated an antibody to a caspase cleavage fragment of actin he calls "Fractin" which stains plaques, dystrophic neurites and microglia (see Am J Pathol, Feb '98). He showed several examples of apoptotic blebbs coming off of dystrophic neurites and synaptic clusters and synapses degenerating and being phagocytosed by microglia. He proposed the term "synapoptosis" for this stripping away of synapses. In double and triple confocal imaging Fractin labelling was sometimes co-localized with APP and AT8, but the majority was independent in both the AD brain and Hsiao mouse.

There was also co-localization with synaptophysin. Cole also reported a decrease in synaptophysin staining around amyloid plaques in entorhinal cortex of the Hsiao mouse. Quantification via Western blot indicates a strong trend for decreased synaptophysin in 12 month vs. 2 month mice, but due to an outlier, the difference is not yet significant. (Hsiao pointed out that Irizarry's report of no synaptophysin decreases in the molecular layer of the dentate gyrus not the entorhinal cortex). In the AD brain, whole microglia around plaques were Fractin immunopositive while positive microglia within the Hsiao mouse were almost never seen. Cole hypothesized that when reactive oxygen species damage mitochondria within the synapse, the release of cytochrome C activates a caspase cascade and degradation of the synapse. The question remains, however, why there is little neuron loss in the Tg mouse compared to man and Cole's suggestion is that there is more caspase labeling in terminals and far less activation of caspase in the microglia of Tg mice. Someone asked if Cole was suggesting that the synapoptosis was presynaptic and not postsynaptic. He replied that he has looked for co-localization of MAP-5 and Fractin and sees little evidence of co-localization. However he also pointed out that his Fractin antibody may not recognize epitopes within dendrites as well as those within axon terminals.