We have previously reported a large Danish pedigree with autosomal dominant frontotemporal dementia (FTD) linked to chromosome 3 (FTD3). Here we identify a mutation in CHMP2B, encoding a component of the endosomal ESCRTIII complex, and show that it results in aberrant mRNA splicing in tissue samples from affected members of this family. We also describe an additional missense mutation in an unrelated individual with FTD. Aberration in the endosomal ESCRTIII complex may result in FTD and neurodegenerative disease.

A significant minority of degenerative dementias lack distinctive inclusion bodies, plagues or tangles on pathological examination. Half of these cases have a positive family history of dementia. We have studied the largest published family with such a dementia and mapped the disease locus to a 12 cM region of chromosome 3 spanning the centromere. Haplotype analysis demonstrates a common region shared between all affected individuals between the markers D3S1284 and D3S1603. Like a number of other late onset neurodegenerative diseases, the disease presents at an earlier age when paternally inherited.

FTD-3 shares clinical and pathologic features with other forms of FTD and fulfills international consensus criteria for FTD. There is involvement of the parietal lobes clinically, radiologically, and pathologically in FTD-3 in contrast to some forms of FTD. This more diffuse involvement of the cerebral cortex leads to a distinctive, global pattern of reduced blood flow on PET scanning.

We have previously localized a locus causing familial nonspecific dementia to the centromeric region of chromosome 3 in a pedigree from the Jutland area of Denmark. This pedigree shows anticipation. Here we present further analysis of these anticipation data which are suggestive of trinucleotide repeat expansion involvement. We also outline our strategies to clone the mutant gene via its putative associated trinucleotide repeat sequence.

We have studied a family in which 14 persons among 73 are or have been suffering from presenile dementia. Post mortem examination showed atrophy but no sign of any known demential syndrome. Cerebral blood flow measured in the late stage of disease was low, but with no characteristic pattern in flow distribution. In one patient in the initial stage of disease, the cerebral blood flow was unexpectedly increased. The patients with presenile dementia in this family did not reveal pathological signs of any known demential syndrome and showed CBF-changes not earlier reported. Moreover, contrary to widely held views we have evidence that dementia may be connected to a high blood flow at least in the initial state. An increased blood flow was also seen in seven of ten well functioning first degree relatives, in some cases along with cerebral atrophy and/or psychological tests with signs of dementia. Are these people going to develop manifest dementia later in life?