Chemical Biology, Molecular Mechanism and Clinical Perspective of γ-Secretase Modulators in Alzheimer's Disease

Abstract

Comprehensive evidence supports that oligomerization and accumulation of amyloidogenic Aβ42 peptides in brain is crucial in the pathogenesis of both familial and sporadic forms of Alzheimer's disease. Imaging studies indicate that the buildup of Aβ begins many years before the onset of clinical symptoms, and that subsequent neurodegeneration and cognitive decline may proceed independently of Aβ. This implies the necessity for early intervention in cognitively normal individuals with therapeutic strategies that prioritize safety. The aspartyl protease γ-secretase catalyses the last step in the cellular generation of Aβ42 peptides, and is a principal target for anti-amyloidogenic intervention strategies. Due to the essential role of γ-secretase in the NOTCH signaling pathway, overt mechanism-based toxicity has been observed with the first generation of γ-secretase inhibitors, and safety of this approach has been questioned. However, two new classes of small molecules, γ-secretase modulators (GSMs) and NOTCH-sparing γ-secretase inhibitors, have revitalized γ-secretase as a drug target in AD. GSMs are small molecules that cause a product shift from Aβ42 towards shorter and less toxic Ab peptides. Importantly, GSMs spare other physiologically important substrates of the γ-secretase complex like NOTCH. Recently, GSMs with nanomolar potency and favorable in vivo properties have been described. In this review, we summarize the knowledge about the unusual proteolytic activity of γ-secretase, and the chemical biology, molecular mechanisms and clinical perspective of compounds that target the γ-secretase complex, with a particular focus on GSMs.

Keywords: Alzheimer's disease; amyloid-β peptide; gamma-secretase; gamma-secretase modulators.; neurodegeneration.

Fig. (1)

Proteolytic processing of APP and the NOTCH receptor by γ-secretase. The Aβ peptide is derived by sequential proteolysis from APP, a ubiquitously expressed type I transmembrane protein. In the amyloidogenic pathway, APP molecules are first cleaved at the cell surface or in early endosomes by β-secretase (BACE1), a membrane bound aspartyl protease, generating a large, soluble ectodomain, APPs-β, and a membrane-bound fragment, C99, that defines the N-terminus of the Aβ sequence [197]. Subsequent cleavage of C99 by the aspartyl protease γ-secretase approximately in the middle of the TMD generates the C-terminus of the Aβ peptide and releases Aβ from APP. γ- secretase generates Aβ peptides of varying length elongated or truncated at the C-terminus, with peptides ending after 40 and 42 amino acids being the predominant species. In addition to cleavage in the middle of the TMD (γ-cleavage), γ-secretase cleaves close to the cytosolic border of the membrane (ε-cleavage). This cleavage liberates the APP intracellular domain (AICD), which may have a function in transcriptional regulation [42]. The NOTCH receptor is synthesized as a 300 kDa precursor that is cleaved by a furin-like convertase in the trans- Golgi compartment and assembled into a mature heterodimer receptor through non-covalent linkage of the resulting protein fragments (S1 cleavage) [27, 28]. At the cell surface, binding of DSL family ligands (Delta/Jagged) induces shedding of the ectodomain by the metalloprotease ADAM10 (S2 cleavage). The resulting, membrane-bound C-terminal fragment termed NOTCH extracellular truncation (NEXT) undergoes cleavage by γ-secretase at the cytosolic border of the membrane (S3 cleavage), which releases the NOTCH intracellular domain (NICD) into the cytosol. NICD travels to the nucleus where it functions as a transcriptional activator [27, 28]. Interestingly, γ-secretase also cleaves the NEXT fragment in the middle of the TMD at sites topologically similar to the cleavage sites in APP that generate the Aβ peptides. These cleavages (S4 cleavage) generate the Nβ peptides, predominantly Nβ21 and Nβ25, which are released into the extracellular space [113]. In addition to NOTCH receptors, a large number of type-1 membrane proteins including NOTCH receptor ligands, APP homologs, ErbB-4, E- and N-cadherin and CD44 have been identified as γ-secretase substrates. Whereas the physiological relevance of γ-secretase-mediated cleavage events in many of these substrates remains to be clarified, suppression of NOTCH receptor processing and signaling has been shown to result in dramatic phenotypes in a variety of organisms [26].

Fig. (2)

Hypothetical model of substrate interaction and processing by γ-secretase (modified after [74]; see text for details).

Fig. (3)

γ-secretase cleavage sites within the TMDs of APP and the NOTCH1 receptor. Cleavage by β-secretase generates the N-terminus of the Aβ sequence. Subsequently, γ-secretase cleaves within the APP TMD at multiple sites. According to the sequential cleavage model, cleavage occurs initially close to the cytosolic border of the TMD at the ε-site or at the alternative ε-site (γ48), and then in two product lines along opposite surfaces of the helical axis of the substrate APP. Most recently, direct evidence for this model has been provided with the detection of corresponding tripeptides that are released after each sequential cleavage step [68]. This study also described detection of a tetrapeptide that would explain conversion of Aβ42 to Aβ38. Evidence for conversion of Aβ40 to Aβ37 is lacking. Cleavage of NOTCH1 by ADAM10 (S2 cleavage) generates a membrane-bound, C-terminal fragment that is a direct substrate for γ-secretase. Cleavage occurs close to the cytosolic border of the TMD and results in release of the NICD domain (S3 cleavage). In addition, γ-secretase-mediated cleavage events in the middle of the TMD generate two peptides, Nβ21 and Nβ25, with differing C-termini analogous to the Aβ peptides (S4 cleavage) [113]. A similar dual-cleavage mechanism has been proposed for other substrates of γ-secretase [114, 198]. GSMs reduce the generation of Aβ42 and of the corresponding Nβ25 peptides, but spare γ-secretase cleavage at the ε-site and the S3-site and liberation of the AICD and NICD domains. This indicates that the topology of the γ-secretase cleavage sites within the TMD is crucial for the modulation of a given γ- secretase substrate by GSMs.

Fig. (4)

Structures of the active-site directed γ-secretase inhibitor L-685,458 and close analogues.

Fig. (5)

Structure of docking-site binders: α-helical peptide D-10 and the three-amino acids extended analogue D-13.

Fig. (6)

Structures of the allosteric site binder DAPT and derivatives.

Fig. (7)

Structures of aryl sulfonamide-based γ-secretase inhibitors. Some of these compounds display selectivity for APP over NOTCH and have been termed NOTCH-sparing γ-secretase inhibitors.

Fig. (8)

Structures of NOTCH-sparing γ-secretase inhibitors targeting a nucleotide-binding domain.

Fig. (9)

Structures of γ-secretase modulators in the class of NSAIDs. Only few NSAIDs are γ-secretase modulators, and naproxen represents an NSAID without this activity.

Fig. (10)

Substrate binders and the structure of Flurbi-BpB, a flurbiprofen-derived photo-reactive probe.

Fig. (11)

NOTCH-sparing sulfonamides.

Fig. (12)

Ibuprofen-derived γ-secretase modulators.

Fig. (13)

Indomethacin-derived γ-secretase modulators.

Fig. (14)

Non-NSAID γ-secretase modulators.

Fig. (15)

Structure of pyrimidine- and purine-based γ-secretase modulators.

Fig. (16)

Piperidine-based γ-secretase modulators.