Mutations in TM1 affect presenilin-1’s ability to cleave substrates

Allison Ilene Janowitz¹, Anne Brunkan², Alison Goate², Biology Department, Washington University, St. Louis, MO¹,Department of Psychiatry, Washington University School of Medicine²

Less then 1% of all Alzheimer’s disease cases are Familial Alzheimer’s Disease (FAD). The age of onset occurs anywhere between 25-60 years of age. It is a debilitating disease that is caused by abnormal neurofibrillary tangles and neuritic plaques composed of β-amyloid (Aβ in the brain that restrict neuron activity and eventually cause neuron death. The effects are gradual loss of memory, loss of speaking ability, and eventually total dependence on a caregiver. Based on studies performed on families with Alzheimer’s disease, three genes have been identified that play a role in FAD. Mutations in the amyloid precursor protein (APP) gene, found on chromosome 21, and presenilin 1 (PS1) and presenilin 2 (PS2) found on chromosomes 14 and 1, respectively, are believed to have an affect on the formation of Aβ. My project dealt with PS1 and studying the effect of point mutations in residues 95-98, and how these mutations affect PS1’s ability to cleave APP and Notch.

APP is a normal protein found in an individual’s brain within the membranes. It is first snipped to form C99, which is later is snipped again, to form the γ-C-terminal Fragment (CTFγ) and β-amyloid. β-amyloid (Aβ) is usually found with two different numbers of amino acids, Aβ40 and Aβ42. Usually there is 90% Aβ40 and 10% Aβ42. However, PS1 and APP mutations change those percentages, causing more Aβ42. Too much Aβ42 is believed to be toxic and to causes the plaques found in the brains of patients with Alzheimer’s disease. PS1 is the protein responsible for cleaving C99 and forming Aβ40 and Aβ42.

Notch is a different substrate that is processed in a similar way as APP. It is cleaved twice to form Notch Intracellular Domain (NICD). Notch is involved in cell fate determination. Like APP, it is cleaved by PS1, and it can be used to test if PS1 is active and functioning.

Endoproteolysis occurs when PS1 is cleaved to form both full-length PS1 and N-terminal fragment (NTF). It is believed that endoproteolysis is needed for PS1 to be active and able to cleave APP. In certain families with FAD, PS1 has a mutation, _E9, where a section including the normal cleavage site is missing. In this case, with the _E9 mutation, endoproteolysis is not necessary for PS1 to be active. _E9 allows the activity of mutant PS1 to be tested without the concern that mutations are actually affecting endoproteolysis. Thus, there are two ways to test point mutations in PS1, either with _E9 already missing or normal PS1 where the only mutation would be the point mutation. PS1/PS2 knockout mouse cells are used to test PS1’s activity. The knockout cells are from mice that have had all presenilin removed from their cells, and consequently have died. The knockout cells are good to test different types of PS1 because they have no endogenous PS1 and the results show only the effects of the PS1 introduced into the cells.

Aside from the _E9 mutation, one mutation in a family with Alzheimer’s disease is a single point mutation in PS1, transmembrane domain 1 (TM1), in which the valine (V) at residue 96 is changed to phenylalanine (F). I tested V96 and surrounding residues to determine their importance for PS1 activity. After plating cells, I transfected the cells with the mutant PS1 and either C99 or ΔE. The mutants either had one residue changed or all four changed in one sample. Immunoprecipitation and western blots were used to determine the effect of mutations. When all four residues, 95-98, are changed, PS1 does not function at all. By changing individual residues within 95-98 in TM1, it became apparent using both PS1 and _E9 constructs that PS1 was unable to cleave APP C99 when the point mutations affected certain amino acids. When residue 96 in section TM1 is mutated to either an F or Y amino acid, which are both larger than the normal one that is usually there, PS1 is unable to cleave APP C99. The other forms of PS1, with V95F, V97F, A98F, and V96 G, L, and A, looked almost identical to wild type PS. This is true in both cases, _E9 and normal PS1.

These mutations in PS1 were also tested by using Notch. The results were almost identical to the C99 data. There was no NICD when all four residues were mutated. There was no NICD with V96Y and barely detectable amounts with residues V96F and G. All other constructs displayed a similar amount of NICD as wild-type PS1. By determining how these mutations cause PS1 to become inactive, we will develop a better understanding of PS1’s function.