24 Jul 2002

Stockholm. Last July, Thomas Sudhof created a buzz in the Alzheimer’s research community when he reported elegant in vitro work suggesting that the C-terminal fragment of AβPP (APPct), the poorly understood second product of γ-secretase cleavage, could enter the nucleus and change gene expression there (see related news item). Here at the 8th International Conference on Alzheimer’s Disease and Related Disorders in Stockholm, Paul Coleman and his colleagues at the University of Rochester, as well as other labs, presented work carrying this clue further.

On Monday, first author Min Zhu presented a poster describing immunohistochemistry of neurons in selected brain regions of postmortem samples from people with AD. The antibody detected the APPct, but not the amino-terminal end, in the nuclei. Its amount correlated to disease status and to how vulnerable that particular brain area is known to be in AD. In a second experiment, the researchers cloned APPct and its binding partner Fe65 into green fluorescent protein vectors and transfected mammalian cells. The poster showed that both proteins co-localize in the nucleus.

The answer to the obvious question-which genes are affected by APPct-mediated transcription-remains elusive. Finding out which genes are differentially expressed in the cotransfected cells clearly is the next experiment, Coleman said, but previous work by his and other groups already suggests some candidates that could be tested. For example, a paper in press in Neurobiology of Disease details microarray and RT-PCR studies of gene expression differences between brain tissue of normally aged versus AD brains. It shows that genes related to trafficking of synaptic vesicles have reduced expression very early in AD, at stages when genes related to synaptic structure are still expressed at normal levels. An example is dynamin, a protein involved in retrograde vesicle transport. Its RNA and protein levels decrease steeply in early AD, while levels of the structural marker PSD95 are still normal. Intriguingly, this affects only the respective isoforms of these proteins that handle synaptic vesicle traffic, not those dealing with trafficking around the Golgi and related organelles in the cell body, Coleman added.

Today, Stavros Therianos in Coleman’s lab is presenting a poster about another class of potential candidate genes, namely developmentally regulated genes that are reexpressed in AD. The general concept that AD might represent a regression to an earlier stage originated about 20 years ago, when Peter Davies first showed that the Alz50 antibody was immunoreactive with neurons not only in AD but also in very early embryonic stages. Since then, other genes have been found to be re-expressed in early AD, notably those that regulate the cell cycle (see cell cycle live chat). Yesterday, Yan Yang and Karl Herrup of Case Western Reserve University in Cleveland, Ohio, presented data showing expression of several cell-cycle genes that were reexpressed in hippocampal brain samples of people who had died with a diagnosis of mild cognitive impairment but not cognitively normal controls. Today’s poster by Therianos et al. presented a new set of homeobox genes that are normally expressed only in hindbrain development. Candidates aplenty, but the link to APPct is still up for grabs.—Gabrielle Strobel