back to Fifth International Conference

T.-W. Kim,* O. G. Hallmark, W. Pettingell, W. Wasco and R. E. Tanzi Genetics and Aging Unit, Department of Neurology, Massachusetts General Hospital-East and Harvard Medical School, Charlestown, MA 02129, USA

• Enter the seminar

The majority of early-onset familial Alzheimer's disease (FAD) appears to be caused by mutations in two recently identified genes: presenilin 1 (PS1) and presenilin 2 (PS2). These two novel genes, PS1, located on chromosome 14, and PS2, on chromosome 1, are significantly homologous to each other and are members of an evolutionarily conserved gene family. The predicted structures of PS1 and PS2 contain six to nine hydrophobic domains, which produces several large and small hydrophilic loops. Neither the first eighty amino acids nor the single large hydrophilic loop are particularly well conserved, suggesting that these regions impart specificity of function or localization to the PS1 and PS2. The normal biological role(s) of the presenilins and the mechanism(s) by which the FAD-linked mutations exert their effect remains unknown. In the present study, we focused on PS2. To begin investigating normal cellular functions of PS2, we examined biosynthesis and processing of this molecule. For regulated expression of PS2 in human neuronal cells, we have established inducible cell lines expressing either wild-type or mutant forms of PS2 under the tight control of the tetracycline-responsive transactivator. In this system, presence of tetracycline in the culture medium supresses PS2 expression, while its withdrawal results in induction of PS2 expression. Western blot analysis revealed that either N-terminal or C-terminal FLAG epitope-tagged PS2 molecules were visualized as single bands with apparent molecular weights of 52 kDa. In addition to the 52 kDa species, we also observed high molecular weight aggregates and a 20 kDa C-terminal fragment. We examined whether the 20 kDa polypeptide is a stable cellular fragment or a non-specific degradation product. Both pulse-chase experiments and cyclohexamide treatment showed that the 20 kDa fragment was a stable cellular polypeptide and predominantly localized to the detergent insoluble fraction. In contrast, the 52 kDa PS2 product was completely extracted by detergent treatment. Our studies demonstrated that PS2 undergoes proteolytic processing to generate a 20 kDa derivative which is associated with the detergent insoluble fraction in the cell (e.g. cytoskeleton). Interestingly, generation of the 20 kDa fragment from mutant forms of PS2 was increased several fold compared to wild-type. These results raise the possibility that the PS2-derived C-terminal fragment is a byproduct of the normal cellular breakdown of PS2. This fragment may serve as a functional cellular effector and represent a potential marker for PS2-linked FAD neuropathogenesis.