[@aph2003]
[@ref]
[@ref2017]
[@ref2009]
[@refa]
[@refb]
[@refc]
[@idf]
[@page2026]
[@allen]
[@allena]
[@allenb]
[@allenc]
[@brainspan]
[@proteins]
[@adam]
[@amyloidbeta]
[@appprotein]
[@betasecretase]
[@presenilin]
[@refd]
[@refe]
[@qbi]
[@qnz]
[@reff]
[@refg]
| Catalytic Subunit | [Presenilin-1](/proteins/presenilin-1) (PSEN1) or [Presenilin-2](/entities/psen2) (PSEN2) |
| Subunits | Presenilin, Nicastrin, APH-1, PEN-2 |
| UniProt (PSEN1) | P49768 |
| PDB | 5A63, 6IDF, 5FN2 |
| Complex Weight | ~230 kDa |
| TM Domains | 20 transmembrane helices (total) |
| Localization | Plasma membrane, endosomes, ER-Golgi |
| Enzyme Class |
Intramembrane aspartyl protease (GxGD type) |
| Diseases |
[Alzheimer's Disease](/diseases/alzheimers) |
Gamma Secretase Complex (Γ Secretase) is an important component in the neurobiology of neurodegenerative diseases. This page provides detailed information about its structure, function, and role in disease processes.
Gamma-secretase (γ-secretase) is a multi-subunit intramembrane-cleaving protease complex that catalyzes the final proteolytic step in the production of Amyloid-Beta peptides from [amyloid precursor protein (APP[/proteins/app-protein. The complex consists of four essential transmembrane protein subunits: presenilin-1 (PS1 or PS2, the catalytic component), nicastrin (NCT), anterior pharynx-defective 1 (APH-1), and presenilin enhancer 2 (PEN-2) (Edbauer et al., 2003). γ-Secretase is one of the most important drug targets in alzheimers research, though its broad substrate repertoire — cleaving over 150 type I transmembrane proteins including Notch receptors — has made therapeutic modulation exceedingly challenging (Bhatt et al., 2022).
The γ-secretase complex assembles from four obligate subunits with a 1:1:1:1 stoichiometry. Assembly occurs in the endoplasmic reticulum (ER), with the mature, active complex trafficked to the plasma membrane and endosomes.
- Genes: psen1 (chromosome 14q24.2) and psen2 (chromosome 1q42.13)
- Size: ~50 kDa; contains 9 transmembrane domains (TMDs)
- Function: Houses the catalytic aspartyl dyad (Asp257 in TMD6 and Asp385 in TMD7 for PS1) that cleaves substrates within the lipid bilayer. During complex maturation, presenilin undergoes endoproteolytic cleavage between TMD6 and TMD7, generating an N-terminal fragment (NTF) and C-terminal fragment (CTF) that remain associated as the active heterodimer (Bhatt et al., 2022).
- Disease relevance: Over 300 mutations in psen1 and ~50 in psen2 cause familial Alzheimer's Disease, making psen1 the most commonly mutated gene in autosomal dominant AD.
- Gene: NCSTN (chromosome 1q23.2)
- Size: ~130 kDa (heavily glycosylated; core protein ~75 kDa); contains 1 TMD
- Function: Functions as the substrate receptor of the complex. The large extracellular ectodomain (~670 amino acids) recognizes the free N-terminus of substrates generated by ectodomain shedding and positions them for entry into the presenilin active site. The crystal structure of nicastrin ectodomain revealed a fold similar to aminopeptidases, though it lacks enzymatic activity (Bhatt et al., 2014).
- Genes: APH1A (chromosome 1q21.2) and APH1B (chromosome 15q22.2); APH1A has two splice variants (APH1aL, APH1aS)
- Size: ~29 kDa; contains 7 TMDs
- Function: Serves as a scaffolding protein, stabilizing the initial presenilin-nicastrin subcomplex during assembly. APH1A-containing complexes are more abundant in the brain; the APH1 isoform determines which six possible γ-secretase complex variants are formed.
- Gene: PSENEN (chromosome 19q13.12)
- Size: ~12 kDa; contains 2 TMDs (hairpin topology)
- Function: Required for the final maturation step — endoproteolytic cleavage of presenilin. PEN-2 binding triggers presenilin autocleavage and activates the complex. PEN-2 also stabilizes the presenilin NTF/CTF heterodimer in the mature complex.
The atomic-resolution structure of human γ-secretase was first solved by cryo-electron microscopy (cryo-EM) at 3.4 Å resolution (PDB: 5A63), revealing the architecture of the intact complex (Bai et al., 2015):
- Transmembrane region: 20 TMDs arranged in a horseshoe-shaped architecture, forming a central cavity within the membrane where substrate entry and catalysis occur
- Extracellular region: Dominated by the large nicastrin ectodomain, which sits atop the horseshoe like a lid, controlling substrate access
- Active site: The two catalytic aspartates (D257 and D385 in PS1) are positioned ~10 Å apart within the membrane interior, consistent with the aspartyl protease mechanism
Subsequent substrate-bound structures (PDB: 6IDF — γ-secretase with Notch; 5FN2 — with DAPT inhibitor) revealed how substrates enter a lateral gate between TMD2 and TMD6, unfold their transmembrane helix, and present the scissile bond to the catalytic aspartates in a β-strand conformation (Zhou et al., 2019).
The complex exhibits remarkable conformational flexibility, with TMD2 and TMD6 acting as dynamic gatekeepers. This plasticity explains how γ-secretase can accommodate diverse substrates and underlies the variable cleavage positions that determine amyloid-beta peptide length.
¶ Normal Function and Substrates
γ-Secretase is not merely an "amyloid-beta-generating machine" — it is an essential protease that processes a vast repertoire of over 150 type I transmembrane substrates after their ectodomains have been shed by metalloproteases or other sheddases. Key substrates include:
The most critical non-app substrate. γ-Secretase releases the Notch intracellular domain (NICD) from Notch1-4 receptors, which translocates to the nucleus and activates transcription of Hes and Hey family genes. Notch signaling is essential for:
- Stem cell maintenance and differentiation
- Intestinal crypt cell homeostasis
- T-cell maturation in the thymus
- Angiogenesis and vasculogenesis
Sequential cleavage of app-protein C-terminal fragments (C99 and C83) generates amyloid-beta peptides of varying lengths and the app intracellular domain (AICD). The complex performs processive cleavage, initially cutting at the ε-site (Aβ48 or Aβ49), then trimming in ~3-residue steps (Aβ48→45→42→38 or Aβ49→46→43→40) (Takami et al., 2009).
- ErbB4: Signaling in neuronal development
- E-cadherin/N-cadherin: Cell adhesion regulation
- CD44: Cell migration and immune signaling
- EphB2: Synaptic plasticity
- lrp1: Cholesterol metabolism and amyloid-beta clearance
γ-Secretase is central to alzheimers pathogenesis through two mechanisms:
-
Aβ generation: The enzyme produces the toxic Amyloid-Beta peptide from app-protein-derived C99 substrate. The ratio of Aβ42/Aβ40 is a critical determinant of amyloid-aggregation and plaque formation.
-
Familial AD mutations: Over 300 psen1 mutations and ~50 psen2 mutations cause familial Alzheimer's Disease. Most FAD mutations do not simply increase total Aβ production; rather, they impair the processive trimming activity of γ-secretase, causing premature release of longer, more aggregation-prone Aβ42 and Aβ43 peptides while reducing production of shorter Aβ38 and Aβ40 species (Szaruga et al., 2017). This "loss-of-function for processivity" model explains why psen1 mutations increase the Aβ42/Aβ40 ratio.
Gain-of-function mutations in Notch pathway components (including γ-secretase subunits) have been identified in T-cell acute lymphoblastic leukemia (T-ALL), making γ potential cancer therapeutics.
Several GSIs advanced to clinical trials for Alzheimer's Disease but all failed:
- Semagacestat (LY450139): Eli Lilly's phase III trial (IDENTITY) was terminated in 2010 after patients showed cognitive worsening and increased skin cancer risk. Post-hoc analysis revealed semagacestat broadly inhibited all γ-secretase substrates, particularly Notch, causing gastrointestinal toxicity, immunosuppression, and paradoxically increased Aβ production due to substrate accumulation during drug troughs (De Strooper, 2014).
- Avagacestat (BMS-708163): Bristol-Myers Squibb's phase II trial stopped due to dose-dependent gastrointestinal and dermatological adverse events, consistent with Notch inhibition. Despite claims of app-selective inhibition, cell-based assays showed inadequate Notch sparing.
- Tarenflurbil (R-flurbiprofen): A γ-secretase modulator that showed substrate selectivity in preclinical models but failed in phase III (2008) due to poor CNS penetration and insufficient Aβ42 reduction.
GSMs represent a more refined approach: rather than blocking all cleavage, they allosterically shift the cleavage position to favor production of shorter, less toxic Aβ species (Aβ37, Aβ38) while reducing Aβ42, without affecting Notch processing. Second-generation GSMs with improved potency and brain penetrance are in preclinical and early clinical development.
Emerging approaches include antisense oligonucleotides and CRISPR-based strategies to correct specific PSEN1 mutations in familial AD patients, preserving normal γ-secretase function while eliminating the pathogenic allele.
The γ-secretase cleavage reaction represents a unique type of proteolysis that occurs within the hydrophobic environment of the lipid bilayer. Unlike classical proteases that cleave in aqueous environments, γ-secretase must recognize and cleave transmembrane substrates that are embedded in the membrane.
The catalytic mechanism involves two essential aspartyl residues within the transmembrane domains of presenilin:
- Asp257 (PS1 TMD6): The primary catalytic aspartate
- Asp385 (PS1 TMD7): The partner in the catalytic dyad
These residues coordinate a water molecule that performs nucleophilic attack on the peptide bond. The mechanism is similar to that of classical aspartyl proteases like pepsin, but occurs within the membrane. The catalytic water is activated through a network of hydrogen bonds involving the two aspartates and surrounding residues, allowing the hydrolysis of the peptide bond while the substrate remains embedded in the membrane (Shen & Kobilka, 2012).
¶ Processive Cleavage and Aβ Species Generation
One of the most distinctive features of γ-secretase is its "processive cleavage" mechanism. Rather than making a single cut, the enzyme performs sequential proteolysis:
- Initial cleavage (ε-site): The first cut occurs at the ε-cleavage site, producing Aβ49 or Aβ48
- Trimming (γ-site): Subsequent cuts occur at 3-residue intervals: Aβ48→Aβ45→Aβ42→Aβ38 or Aβ49→Aβ46→Aβ43→Aβ40
The final Aβ species produced depends on the efficiency of these subsequent trimming steps. Mutations that impair processivity cause premature release of longer, more aggregation-prone Aβ42/43 species (Szaruga et al., 2017).
¶ Substrate Recognition and Entry
Substrate entry occurs through a lateral gate between TMD2 and TMD6 of presenilin. Key features of substrate recognition include:
- Ectodomain shedding requirement: Substrates must first undergo proteolytic removal of their extracellular domain by sheddases (α-, β-secretases)
- Nicastrin ectodomain: The large extracellular domain of nicastrin acts as a "docking station" for the N-terminus of the substrate after shedding
- Transmembrane helix unfolding: Substrates must partially unfold their transmembrane helix to adopt a β-strand conformation for presentation to the active site
- Lateral gate dynamics: TMD2 and TMD6 undergo conformational changes to allow substrate entry and release of products
¶ Assembly and Cellular Trafficking
The γ-secretase complex assembles through an ordered, stepwise process in the endoplasmic reticulum:
- Initial dimerization: Presenilin forms dimers (or requires NCT for stable expression)
- NCT binding: Nicastrin associates with presenilin to form the core subcomplex
- APH-1 incorporation: APH-1 stabilizes the PSEN-NCT intermediate
- PEN-2 addition: PEN-2 binds last, triggering presenilin endoproteolysis
- Maturation: The complex undergoes glycosylation and traffics through the Golgi
Active γ-secretase is found in multiple cellular compartments:
- Plasma membrane: Primary location for APP processing
- Endosomes: acidic pH optimizes γ-secretase activity; endosomal trafficking is critical for amyloidogenic processing
- ER-Golgi intermediate compartments: Some processing occurs here
- Lipid rafts: Membrane microdomains enriched in cholesterol and sphingolipids concentrate γ-secretase activity
The subcellular distribution of γ-secretase determines which substrates are preferentially processed and influences the products generated.
¶ Inhibition Strategies and Challenges
The early approach to AD therapy targeted γ-secretase directly:
- Transition state mimetics: Designed to mimic the tetrahedral intermediate of peptide bond hydrolysis
- Notch-sparing inhibitors: Attempted to achieve substrate selectivity
- Allosteric modulators: Target regulatory sites rather than the active site
The fundamental challenge is that γ-secretase's active site is a "promiscuous" pocket evolved to accommodate diverse substrates. Achieving selectivity for APP over Notch has proven extremely difficult.
The failures of semagacestat and other GSIs in clinical trials taught crucial lessons:
- Notch inhibition is dose-limiting: Even "Notch-sparing" inhibitors cause sufficient Notch inhibition at therapeutic doses to cause side effects
- Aβ rebound: With partial inhibition, substrate accumulation during drug troughs can paradoxically increase Aβ production
- Broad substrate inhibition: γ-secretase processes many proteins essential for normal neuronal function
- Ineffectiveness in advanced disease: Trials enrolled patients with already-established pathology
GSMs represent a fundamentally different approach:
- Mechanism: Allosteric modulation shifts cleavage position rather than blocking activity
- Effect: Increased production of shorter Aβ37/Aβ38, decreased Aβ42
- Advantage: Preserve Notch processing and other essential functions
- Challenge: GSMs are typically small molecules with limited brain penetration
Second-generation GSMs with improved potency and brain availability are in development. Some show disease-modifying potential in animal models.
¶ Genetic Variation and Disease
Over 300 PSEN1 and ~50 PSEN2 mutations cause autosomal dominant AD. These mutations have provided crucial insights:
- Pathogenic mechanism: Most FAD mutations alter Aβ42/Aβ40 ratio rather than total Aβ production
- Loss-of-function aspect: Some mutations reduce γ-secretase activity, contributing to synaptic dysfunction
- Penetrance: Nearly 100% of carriers develop AD, typically with onset 30-60 years
- Neuropathology: PSEN1 FAD shows earlier onset, often with spastic paraparesis or atypical features
Beyond rare deterministic mutations, common variants in γ-secretase subunit genes influence AD risk:
- NCSTN variants: Some polymorphisms associated with altered AD risk
- APH1B variants: Linked to lipid metabolism and AD in genome-wide studies
- PSENEN variants: Associated with immune function and neurodegeneration
- Cell-free assays: Purified reconstituted γ-secretase for mechanistic studies
- Cell-based systems: Overexpression of APP and γ-secretase components
- iPSC-derived neurons: Human disease modeling with patient mutations
- Organoids: 3D brain models for drug testing
- APP/PS1 mice: Commonly used for Aβ pathology studies
- PSEN knock-in models: Express FAD mutations in endogenous loci
- Conditional knockouts: Tissue-specific deletion to study function
- Humanized mice: Express human APP and γ-secretase for translational studies
Rather than inhibiting γ-secretase directly, emerging approaches target specific substrates:
- APP-selective inhibitors: Small molecules designed to preferentially reduce APP cleavage
- Aβ42 antibodies: Passive immunization to neutralize Aβ42 specifically
- BACE inhibitors in combination: Reduce Aβ production upstream of γ-secretase
- ASO targeting PSEN1: Antisense oligonucleotides to reduce PSEN1 expression
- CRISPR-based editing: Correct specific mutations in patient-derived cells
- RNAi approaches: Reduce toxic PSEN1 variants while preserving wild-type
- Autophagy enhancers: Improve clearance of Aβ aggregates
- BBB-penetrant compounds: Improve drug delivery to CNS
- Combination approaches: GSMs + immunotherapies
The study of Gamma Secretase Complex (Γ Secretase) has evolved significantly over the past decades. Research in this area has revealed important insights into the underlying mechanisms of neurodegeneration and continues to drive therapeutic development.
Historical context and key discoveries in this field have shaped our current understanding and will continue to guide future research directions.
The advent of cryo-electron microscopy (cryo-EM) revolutionized our understanding of γ-secretase structure. The 2015 publication of the first atomic-resolution structure (PDB: 5A63) at 3.4 Å revealed the complete architecture of the human γ-secretase complex (Bai et al., 2015).
Key structural features include:
- Horseshoe-shaped transmembrane domain: The 20 transmembrane helices form a U-shaped structure with the active site at the base
- Nicastrin "cap": The large extracellular domain of nicastrin sits atop the transmembrane region like a lid
- Internal cavity: A large internal cavity within the membrane serves as the proteolytic chamber
- Lateral gate: The interface between TMD2 and TMD6 provides the entry point for substrates
Subsequent structures with bound substrates and inhibitors (PDB: 6IDF, 5FN2) revealed the conformational changes required for substrate binding and catalysis (Zhou et al., 2019).
The catalytic site consists of:
- Asp257 in TMD6: Part of the conserved DTL motif
- Asp385 in TMD7: Part of the conserved DTL motif
- Water-bridge network: Structured water molecules that mediate catalysis
- Surrounding residues: Hydrophobic and polar residues that recognize substrate backbone
The active site is positioned approximately 10 Å from the membrane surface, allowing access to transmembrane portions of substrates while maintaining the aqueous environment required for hydrolysis.
Cryo-EM studies have revealed multiple conformations:
- Apo state: The enzyme adopts an "open" conformation when no substrate is bound
- Substrate-bound state: The lateral gate closes around the substrate
- Inhibitor-bound state: Various inhibitor classes induce distinct conformational changes
- Translocation intermediates: Images capture the progressive movement of substrates through the cleavage cycle
This flexibility is central to understanding how γ-secretase accommodates its diverse substrate repertoire.
¶ Notch Signaling and Essential Functions
Beyond APP processing, γ-secretase is essential for Notch receptor signaling:
Notch Processing Pathway:
- Notch ligands (Delta, Jagged) bind to Notch receptors
- Proteolytic cleavage (S1) by ADAM proteases releases the Notch extracellular domain
- γ-secretase performs S2 cleavage, releasing the Notch intracellular domain (NICD)
- NICD translocates to the nucleus and activates transcription
Notch Functions in the CNS:
- Neural stem cell maintenance and differentiation
- Synaptic plasticity and memory formation
- Oligodendrocyte development and myelination
- Astrocyte differentiation
The essential nature of Notch signaling explains why complete γ-secretase inhibition is toxic.
¶ Synaptic Function and Neuronal Activity
γ-secretase activity affects synaptic function through multiple mechanisms:
- Aβ production at synapses: Synaptic activity modulates γ-secretase localization and activity
- Notch-dependent plasticity: NICD signaling affects long-term potentiation (LTP)
- Other substrates: CD44, Eph receptors, and adhesion molecules modulate spine morphology
This intersection of γ-secretase biology with synaptic function is a key area for understanding how γ-secretase modulation might improve or worsen cognitive outcomes.
¶ Trafficking and Endocytic Recycling
The subcellular distribution of γ-secretase critically affects its function:
- Endosomal processing: Acidic endosomal pH optimizes γ-secretase activity
- Synaptic vesicles: Some γ-secretase localizes to presynaptic compartments
- Recycling endosomes: APP and γ-secretase recycle together
- Axonal and dendritic trafficking: Differential distribution affects processing in different neuronal compartments
Understanding trafficking has led to strategies targeting endosomal γ-secretase as a more selective approach.
¶ Biomarkers and Diagnostic Applications
The Aβ42/Aβ40 ratio measured in CSF is a key AD biomarker:
- Aβ42 decline: Reflects plaque formation (Aβ42 gets "trapped" in plaques)
- Aβ40 reduction: Less specific to AD
- Ratio utility: Aβ42/Aβ40 provides better discrimination than either species alone
Direct measurements of γ-secretase activity are challenging but important:
- Aβ production rates: Measured in brain slices or CSF
- Notch cleavage products: NICD levels in accessible tissues
- Substrate accumulation: During pharmacological inhibition
These biomarkers could help identify optimal dosing for γ-secretase modulators.
PSEN1 and PSEN2 genetic testing is available:
- Diagnostic testing: For early-onset AD with family history
- Predictive testing: For at-risk individuals (controversial)
- Clinical trials: Pre-screening for specific mutations
¶ Future Directions and Outstanding Questions
- Why do GSIs fail clinically? The disconnect between biomarker effects and clinical outcomes
- How do GSMs work? The allosteric mechanism remains incompletely understood
- Can substrate selectivity be achieved? Targeting APP without affecting Notch
- What is the normal physiological role of Aβ? Could provide insight into therapeutic approaches
- Single-molecule studies: Understanding stochastic cleavage events
- Patient-derived models: iPSC neurons with FAD mutations
- Structural dynamics: Time-resolved cryo-EM to capture intermediates
- Computational approaches: Machine learning for inhibitor design
- Edbauer et al., Aph-1, Pen-2, and Nicastrin with Presenilin generate an active γ-secretase complex (2003)
- Bai et al., An atomic structure of human γ-secretase (2015)
- Bhatt et al., γ-Secretase in Alzheimer's Disease (2022)
- Zhou et al., Structure and function of γ-secretase (2019)
- Szaruga et al., Qualitative changes in γ-secretase activity (2017)
- De Strooper, Lessons from a failed γ-secretase Alzheimer trial (2014)
- Shen & Kobilka, GPCR structure and mechanism (2012)
[@allen]: - Allen Brain Atlas
[@allena]: - Allen Human Brain Atlas: Gamma-Secretase Complex search
[@allenb]: - Allen Mouse Brain Atlas: Gamma-Secretase Complex search
[@allenc]: - Allen Cell Type Atlas
[@brainspan]: - BrainSpan Developmental Transcriptome## See Also
[@proteins]: - [Proteins Index
[@adam]: - adam10
[@amyloidbeta]: - Amyloid-Beta (Aβ)[/proteins/[Amyloid-Beta
[@appprotein]: - app-protein
[@betasecretase]: - Beta-Secretase 1 ([BACE1/proteins/[bace1-protein
[@presenilin]: - presenilin-1## External Links
[@refd]: - UniProt (PSEN1): P49768
[@refe]: - UniProt (Nicastrin): Q92542
[@qbi]: - UniProt (APH1A): Q96BI3
[@qnz]: - UniProt (PEN-2): Q9NZ42
[@reff]: - PDB: 5A63 (apo structure), 6IDF (Notch-bound), 5FN2 (DAPT-bound)
[@refg]: - AlphaFold (PSEN1): P49768