Notch receptor processing and CNS disease

Abstract

We have studied the relationship of Notch receptor processing and two diseases of the central nervous system: Alzheimer's disease (AD) and Cerebral Autosomal Dominant Arteriopathy with Subcortical Infarcts and Leukoencephalopathy (CADASIL). The Notch receptor signalling pathway plays a crucial role during development in many different species for directing adjacent cells to adopt different fates. During maturation and ligand activation the Notch receptor undergoes a series of complex proteolytic cleavages before the intracellular domain is released and translocated to the nucleus where it affects transcription. The first cleavage (S 1) occurs during the transport out to the plasma membrane in the trans-Golgi compartment by a furin-like convertase. This cleavage is constitutive and generates a bipartite receptor, where the two Notch fragments are held together through non-covalent forces.

The second cleavage (S2) occurs near the transmembrane region at the extracellular side when Notch is ligand- activated. The S2 cleavage is immediately followed by the third cleavage (S3) at the border of the transmembrane region and the intracellular side. The metalloprotease TACE (TNFalpha converting enzyme) has been shown to be responsible for the S2 cleavage, while the S3 cleavage is dependent on presenilins (PSs). The PSs have been genetically linked to familial AD and are involved in the gamma-secretase cleavage of the betaamyloid precursor protein (APP) and the production of the neurotoxic Abeta peptide, which causes senile plaques in the brain. There are four known mammalian Notch receptors (Notch 1-Notch 4) with overlapping patterns of expression, but the role of the different Notch receptors is not that well understood. Mutations in the Notch 3 gene cause the genetic disorder CADASIL. CADASIL is an inherited form of stroke and dementia and the histopathological hallmarks are degeneration of vascular smooth muscle cells and deposits of granular osmiophilic material between the vascular smooth muscle cells. The mutations in the Notch 3 gene always affect a cysteine residue in the epidermal growth factor-like- repeats in the extracellular domain, but how this leads to CADASIL is poorly understood.

There is an obvious need to produce therapeutical drugs that prevent the formation of Abeta peptide, in order to reduce plaque formation in AD, and inhibitors of PSs and the gamma-secretase cleavage would be useful. The fact that not only APP, but also Notch cleavage is controlled by PSs, however complicates the issue, and it will be important to find drugs that act differentially on Notch and APP. We show in a cell-based assay that PSs are absolutely required for S3/gamma-secretase processing of Notch and APP (paper I and II). The assay was further developed and gamma-secreatse cleavage of Notch and APP could be measured by adding different gamma-secretase inhibitors and PS active site mutants. We demonstrate that the assay is specific and sensitive and can record small differences in the gamma-secretase processing of Notch and APP (paper II). Furthermore, in an in vitro assay, we show that PSs are not the direct executors of Notch processing, but that the enzymatic activity is absolutely dependent on PSs. This notion is based on the observation that there is a specific Notch-processing activity in PS-containing cells, which is absent in PS-deficient cells, but the activity does not comigrate with PSs in protein purification (paper II). To learn more in detail which parts of the PSI protein are crucial and responsible for gamma-secretase cleavage of Notch and APP we set out to narrow down the PSI protein by deletions and mutations (paper III). We show that a five amino acid motif in the N-terminal fragment of PS I is necessary for gamma-secretase cleavage of both Notch and APP. Moreover, a tyrosine residue in position 288 is critical and the activity is restored when substituting this amino acid with hydrophobic or bulky, but not with charged residues. The endoproteolysis of PS I is in some but not all cases affected when different residues occupy the 288 position, indicating a more direct role of the Tyrosine 288 in the gamma-secretase activity.

To address the cellular and biochemical mechanisms behind CADASIL, we have used a very prevalent mutation in the Notch 3 gene (R142C), which cause CADASIL in humans (paper IV). Using coculture experiments, western blot analysis and immunocytochemistry of stable cell lines expressing the CADASIL-mutated or wild type receptor, we show that the S I cleavage is impaired and less amount of the mutated receptor is expressed at the cell surface compared to wild type receptor. Furthermore, the mutated receptor is more prone to form intracellular aggregates, which are found in close vicinity to the Golgi network, but do not share the characteristics of an aggresome. In coculture experiments together with cells expressing the ligands, Serrate and Delta, we find that the Notch 3 receptor can respond to ligand activation but that there are no differences in signalling between the mutated and wild type receptor.

In conclusion, we show that PSs are absolutely required for gamma-secretase cleavage of Notch and APP but that PSs are not the direct executors for Notch cleavage. The cell-based assay allows screening and detection of compounds that differentially affect processing of Notch and APP, which is crucial in the attempt to prevent AD without affecting Notch signalling. We find a critical motif in the PS I gene, tyrosine 288, which seems to affect gamma-secretase cleavage directly for both Notch and APP. Finally, cells expressing as CADASIL-mutated Notch 3 receptor show impaired S I cleavage and decreased expression at the cell surface, more intracellular aggregates but with normal ligand-induced signalling as compared to wild type Notch 3 receptor. These differences can be one of the reasons underlying the cellular mechanisms of CADASIL.