Alzheimer’s disease (AD) is a frequent brain pathology of the elderly, with a far more complicated aetiology than what was thought in the 1990s. The complexity particularly comes from the co-existence of two degenerating processes, tau aggregation and amyloid beta (Aβ) deposition, that affect polymodal association brain areas, a feature never observed in non-human primates and difficult to model.1 Genetic studies have shown that Aβ precursor protein (PP) plays a central role in familial autosomic dominant AD.2 Familial AD (FAD) and sporadic AD (SAD) have the same neuropathological phenotype, with Aβ accumulation in the gray matter of the neocortex.3 Thus, there is a huge body of evidence that Aβ is the neurotoxic that causes both FAD and SAD.4 The question is whether it is that simple. The National Institute for Health and Medical Research (Inserm) Research Center 815 have noted that the basic mechanisms are different in FAD and SAD, as there is a clear overproduction of Aβ species in FAD as well as a modification of the ratio of Aβ of 42 to 40.5 This is not observed in SAD and it is assumed that the aetiology could be linked to a fibrillogenesis or clearance dysfunction of APP metabolites.
Furthermore, in most studies, the role of tau has been understated for a long time. To apprehend this role, the Inserm team has developed a spatio-temporal analysis of tauopathy in many brain areas of hundreds of non-demented and demented patients. This prospective multidisciplinary study showed that tauopathy always progresses in the brain along a precise and invariable pathway, from the entorhinal then hippocampal formation to polymodal association areas and ending in primary regions and many subcortical areas. The cognitive impairment follows the progression of the affected brain regions.6 In strict parallelism, neocortical Aβ deposits increase in quantity and heterogeneity, suggesting a direct link between both neurodegenerative processes.1 Deciphering this link is the key to finding a relevant therapeutic strategy.
SAD – A Tauopathy Fuelled by APP Dysfunction
The parallelism and synergy between tau and Aβ aggregation led the Inserm team to search for an APP molecular event linking the two degenerating processes. APP is a ubiquitous protein found in all cell types of all species, suggesting a basic and important role that remains to be identified. A neurotrophic activity for APP and secreted APP (sAPP) is often mentioned.7 Therefore, a loss of function of APP rather than a gain of toxic function of Aβ could also be a reasonable hypothesis to explain the stimulation of tauopathy and neurodegeneration.
Complementary to this study of Aβ species, the Inserm team found no obvious modification of APP holoprotein in correlation with the pathology. However, carboxy-terminal fragments of APP (APPCTFs) were found to be significantly diminished during the course of AD and well correlated with the progression of tauopathy.8 Beta, alpha and gamma stubs were also significantly decreased in the brain tissue of individuals having an inherited form of AD linked to mutations of presenilin 1, showing a general defect common to FAD and SAD. An important role of the gamma stubs (also named APP intracellular domain (AICD)) as gene regulators could explain their involvement in the disease, as these fragments are dramatically reduced in AD.9–11 These observations led to other therapeutic strategies focused on the concept of a loss of function of APP-stimulating tauopathy, in good agreement with other teams.12–14 Thus, restoring the neuroprotection properties of APP could be a therapeutic strategy to slow down or cure AD.
The ‘Four-hit’ Hypothesis
In the literature, the key events of APP metabolism are generally summarised as follows:
• The neurotoxic Aβ is produced following an N-terminal cleavage by beta-site APP-cleaving enzyme (BACE)1, an aspartic protease that releases a beta stub (see Figure 1).15 This beta stub is then cleaved by a γ-secretase activity to release Aβ and a cytosolic fragment named AICD.16 This is the amyloidogenic, or ‘the evil’, pathway.
• In parallel, an APP cleavage takes place inside the Aβ region through an α-secretase activity (possibly a disintegrin and metalloproteinase domain 10 (ADAM 10) or ADAM 17) releasing an alpha stub. The latter is cleaved by the γ-secretase activity to release p3 and AICD.17–19 This is the non-amyloidogenic, or ‘the good’, pathway, as there is no production of the neurotoxic Aβ but a release of metabolites with neurotrophic properties – sAPPα and AICD.
Stimulating the non-amyloidogenic pathway at the expense of the amyloidogenic one seems the most promising avenue to prevent or treat AD.20
BACE1 As a Therapeutic Target
BACE1 is, in principle, an excellent therapeutic target for strategies to reduce the production of Aβ in AD. However, it is known that BACE1 also has important physiological roles and that knocking down this enzyme is lethal.21 The beta stubs are normal metabolites of the brain and their total suppression could be detrimental. They are also already decreased in AD.22 These data challenge the general idea of BACE1 as a safe drug target.21 Interestingly, however, a moderate decrease in BACE1 activity provokes a shift towards the non-amyloidogenic pathway. This has been observed for BACE inhibitors such as 4-(2- aminoethyl)benzene-sulphonylfluoride (AEBSF).23 The possible inhibition of γ-secretase to block Aβ secretion has also been explored, but this secretase activity is involved in ubiquitous and important physiological pathways rendering this approach dangerous.24–26
-Secretase As a Therapeutic Target
Another possibility is to stimulate the α-secretase pathway, which is generally activated through stimulation of muscarinic receptors or protein kinase C (PKC) activation.27–30 Drugs such as epigallocatechin gallate (EGCG) have successfully elevated α-secretase activity and reduced the number of plaques on a transgenic model of amyloidosis in AD.31,32
Aβ As a Therapeutic Target
Aβ could be the neurotoxic that causes AD. However, if true, the neurotoxic mechanisms involved are not yet known; it is uncertain whether it is through plaques, protofibrils or oligomers, and it could be when this peptide is either inside or outside the cell. It is unclear whether the N-truncated forms are more toxic than full-length Aβ; they may be toxic at the first stage of the disease or in later stages.
Whatever the answers, there is consensus that anti- Alzheimer drugs should decrease, maybe moderately, the secretion of Aβ peptide, something which both BACE1 inhibition and α-secretase activation do.
Tau As a Therapeutic Target
Tau is a therapeutic target for more than 15 degenerative disorders.33 However, the Inserm team has noted that the sequential pattern of tauopathy in AD, along cortico-cortical connections, is likely to be explained by a loss of neurotrophic factor. Theoretically, as nerve cell populations survive through a chemical cross-talk of neurotrophic factors, if one brain area is degenerating, a progressive collapse of the neuronal network is expected. Stimulating the production of sAPPα and AICD while inhibiting the possible neurotoxic factor should, hypothetically, slow down the progression of tauopathy in AD.
From its study on tau and Aβ in the human brain, the Inserm team proposes that a good anti-Alzheimer drug should increase the α-secretase activity and thus:
• decrease the production of beta stubs and Aβ, a potential neurotoxic;
• increase, in the same mechanism, the secretion of sAPPα, a potential neurotrophic factor;
• increase the production of the potential transcription factor AICD; and subsequently
• the beneficial effects should slow down the progression of tauopathy.
Theoretically, this drug should be able to reduce or stop the deleterious effect of AβPP loss of function, and, thus, be able to stop the burden that fuels tauopathy and provoke dementia in AD. As already mentioned, drugs that have this property have already been developed.32,34 ■