Dermaut works with Christine van Broeckhoven, whose lab tracks inherited dementias in adults and has uncovered great genetic and phenotypic heterogeneity among them, as have other groups. In Philadelphia, Dermaut related the story of a man who at age 50 lost initiative, became apathetic, emotionally blunt, and showed disinhibition of the frontal cortex (a polite scientific way of saying people behave embarrassingly and impulsively in public)—all tell-tale clinical signs of frontotemporal dementia (FTD). Brain imaging at that stage revealed atrophy in frontal and temporal regions of his cortex. At 59, the man’s cognition deteriorated severely; he died at 62. The gross pathology of his brain (dramatic atrophy, thinning of gyruses, ventricular enlargement) fit the pattern of FTD, and the molecular pathology (tau-positive Pick bodies and balloon cells in cortex) then led to a postmortem diagnosis of Pick’s disease as a subtype of the FTD complex. So far, clinic and pathology fit neatly into a little box.

But the genetics changed that. The man’s family history showed that his father had a slowly progressive cognitive deterioration, and a brother had had a personality disorder and taken his own life. Where one would have expected to find a tau mutation in this family, the researchers instead found a presenilin1 mutation. It segregated in the living family members in ways suggesting it is the right one. Of six younger siblings, some have the mutation but not (yet?) any symptoms. Though their clinical follow-up is still ongoing, Dermaut has noted mild atrophy in brain scans of some carriers, one of whom has clinical dementia. “We think this PS1 mutation causes this Pick version of tauopathy,” says Dermaut.

What is different about this PS1 mutation? Why did this man not develop AD? His pathology showed no amyloid plaques. The usual increase in the 42/40 ratio that most PS1 mutations cause appeared not to be there, either, though the presented data did not definitively answer this question. Closer inspection of this new Gly183Val mutation (Dermaut et al., 2004) yielded a clue, however. It alters a conserved base in the splicing signal at the sixth exon and leads to a splicing defect that leaves the PS1 protein without exons 6, or 6 and 7. This suggests that this heterogeneous form of neurodegeneration could be a splicing disease. Scientists Down Under have already found prior examples of this (see Evin et al., 2002). More broadly, several splice errors have been reported for some forms of AD, and they keep cropping up increasingly in other disease fields, as well.

“This short splice form of PS1 induces a non-amyloidogenic pathway that leads to a tauopathy,” said Dermaut. How would this happen? Dermaut had no answer but suggested that the mechanism may be related to presenilin loss of function. Samir Kumar-Singh, a member of van Broeckhoven's team, gave a separate talk about this unusual family, in which he pointed out that he found intracellular Aβ reactivity in autopsy brain tissue of a family member. Kumar-Singh analyzed the presence of tau in this brain and suggested that this PS1 mutation might interfere, directly or indirectly, with tau phophorylation or aggregation states. Furthermore, Dermaut pointed to a recent study, where deleting presenilin function in the cortex of mice led to tau hyperphosphorylation and neurodegeneration (see Saura et al., 2004 in ARF related news story). Beyond this published work, numerous disparate presentations in Philadelphia pointed to a loss of function for presenilin. As always, we invite our readers to fill the gaps in this story.—Gabrielle Strobel.