The Alzheimer’s disease (AD) brain displays extracellular plaques of amyloid-β (Aβ), neurofibrillary tangles of tau and neuronal loss. The 40-42 amino acid Aβ peptide is formed from the amyloid precursor protein (APP) by β-secretase and γ-secretase, while α-secretase prevents Aβ generation. According to the amyloid cascade hypothesis, AD is initiated by increased brain levels of toxic Aβ species. Our laboratory has previously identified two APP mutations, causing early onset AD. Whereas the Swedish mutation results in increased β-secretase cleavage leading to higher Aβ levels, the Arctic mutation leads to a conformational Aβ change that promotes formation of toxic Aβ protofibrils.

In this thesis, we have performed a screen for novel disease-causing mutations in 102 patients with early onset dementia disorders, who underwent investigation at the Memory clinic at Uppsala University Hospital. Mainly, we found a new APP mutation, which causes familial dominantly inherited AD with age at symptom onset in the early forties. This Uppsala APP mutation, consists of an intra-Aβ deletion of six consecutive amino acids, which results in Aβ with 34-36 amino acids (AβUpp_Δ19-24).

Affected mutation-carriers develop symptoms typical of AD. As for biomarkers, the patients display expected changes although brain Aβ imaging by [¹¹C]PIB-PET is only slightly pathological and Aβ42-analysis of cerebrospinal fluid yields normal results. By investigating neuropathological, biochemical and structural properties of AβUpp_Δ19-24 in patient samples, on synthetic peptides and in cell culture models we found evidence that Uppsala APP is pathogenic via three mechanisms: increased β-secretase cleavage, altered α-secretase cleavage and rapid formation of Aβ fibrils into unique polymorphs.

To allow for in vivo studies of molecular mechanisms related to the Uppsala APP mutation we generated transgenic mice, expressing human APP with this mutation together with Swedish APP (to increase Aβ levels). In the brain of tg-UppSwe mice, we observed diffuse aggregates of mainly AβUpp42_Δ19-24, which, given their normal γ-secretase activity, distinguishes these mice from most transgenic mouse models. In order to study if AβUpp_Δ19-24 co-aggregates with wild-type Aβ (Aβwt), we crossed tg-UppSwe with tg-Swe. Analyses of brains from such mice indicated that AβUpp_Δ19-24 may act as seeds for Aβwt by changing its aggregation behavior and thereby increasing its deposition in brain.

Taken together, our studies of the Uppsala APP mutation have provided new knowledge of pathogenic molecular mechanisms in AD and of basic Aβ biology. Such insights may in a longer perspective help us to develop new diagnostics and therapeutics for this disorder.