Parkin is associated to autosomal recessive early-onset Parkinson's disease. Parkin acts as an E3-ubiquitin ligase involved in the proteasome-mediated degradation of various substrates. It has been suggested that pathogenic mutations of parkin, abolishing its ubiquitin-ligase activity, could explain the accumulation of proteins and lead to neuronal death by apoptosis. However, besides this function, additional parkin-dependent cellular pathways exist. We demonstrated that parkin is a direct transcriptional repressor of the tumor suppressor p53 1 . p53 regulates the expression and functions of presenilin-1 (PS1) and presenilin-2 (PS2), two members of the gamma secretase complex involved in the production of the amyloid β peptide (Aβ) and parkin could control the homeostasis of intracellular Aβ. These findings prompted us to investigate whether parkin could control presenilins and if so, whether it is via a direct transcriptional control of PS promoters or indirectly, via p53.

Experiments were conducted on TSM1 neurons, SH-SY5Y human neuroblastoma cells, human embryonic kidney 293 cells, and immortalized mouse embryonic fibroblasts invalidated or not for parkin, presenilin 1 and/or 2, p19^arf and both p19^arf p53. We also used primary cultured neurons and brain extract from mouse invalidated or not for parkin.

We did Q-PCR, Western-blot, caspases-3 activity measurement and in vitro gamma secretase assays experiments. We document by chromosome immune-precipitation, gel shift, gene reporter and mutagenesis experiments parkin direct interaction with presenilins promoters.

Parkin controls presenilin 1 and 2 expressions, promoter activity, and mRNA levels ex vivo and in mouse brains. This regulation impacts on PS-dependent γ-secretase activity and presenilin-mediated control of cell death. This control is independent of parkin ubiquitin-ligase activity, does not involve p53 and is not affected by PS1 and PS2 functional interplay. Parkin binds to presenilins promoters via a consensus binding sequence that we identify and validate by functional analysis 2 .

This study is a "framework" for the identification of novel transcriptional targets of parkin and for a better comprehension of parkin's functions.