PSEN2

PSEN2 encodes presenilin-2, the catalytic subunit of one of two distinct γ-secretase complexes that cleave type I transmembrane substrates including APP and Notch receptors. A unique cytoplasmic motif in PSEN2 directs its γ-secretase complex to late endosomes/lysosomes via phosphorylation-dependent interaction with the AP-1 adaptor complex, restricting substrate cleavage to acidic compartments and generating an intracellular Aβ pool enriched in longer Aβ42 species—an effect amplified by familial Alzheimer's disease (FAD) mutations such as N141I (PMID:27293189, PMID:29078805). Independent of γ-secretase activity, PSEN2 modulates ER calcium stores, and FAD mutations partially deplete ER Ca²⁺, impairing autophagosome–lysosome fusion through reduced RAB7 recruitment (PMID:30892128, PMID:29945658). PSEN2 also supports Notch-dependent neurogenesis and neural crest development (PMID:19563801), selectively restores PDGF receptor expression and PI3K/Akt–ERK signaling through an N-terminal fragment–dependent, γ-secretase-independent mechanism involving the co-activator FHL2 (PMID:16014629), and when carrying FAD mutations enhances neuronal apoptosis via Bcl-2 downregulation and exaggerates microglial inflammatory responses (PMID:11752057, PMID:32741831). Mutations in PSEN2 cause familial Alzheimer's disease, as demonstrated by isogenic CRISPR-corrected iPSC models that normalize both elevated Aβ42/40 ratios and electrophysiological deficits (PMID:29078805).

1. 1996 Medium

   Establishing the genomic architecture and expression profile of PSEN2 provided the foundation for studying its tissue-specific roles and splice regulation.

   Evidence Genomic sequencing and Northern blot in human tissues

   PMID:8661049

   Open questions at the time

   • Functional significance of alternative splicing at exon 10 was not determined
   • Protein-level expression across tissues was not characterized
2. 1998 Medium

   Demonstrating that full-length mutant PS2 (both N- and C-terminal fragments cooperating) is required for Aβ42 overproduction established PS2 as an integral part of the amyloidogenic cleavage machinery rather than a passive bystander.

   Evidence Truncation mutagenesis of PS2 with Aβ42 ELISA in COS/neuro2a cells; complementary study showing WT PS2 promotes α-secretase cleavage while N141I suppresses it

   PMID:9694871 PMID:9813158

   Open questions at the time

   • Mechanism by which N- and C-terminal fragments cooperate was not resolved
   • Whether PS2 directly cleaves APP or acts indirectly was still unknown
3. 2001 Medium

   Linking PS2 overexpression (especially N141I) to caspase-3-dependent apoptosis and Bcl-2 downregulation in primary neurons revealed a pro-apoptotic role potentially contributing to neurodegeneration beyond Aβ production.

   Evidence Adenoviral transduction of PS2 in rat cortical neurons, caspase-3 inhibition, Bcl-2 Western blot

   PMID:11752057

   Open questions at the time

   • Whether Bcl-2 downregulation is transcriptional or post-translational was not determined
   • Relevance in vivo at endogenous expression levels was not tested
4. 2004 Medium

   Showing that mutant PS2 promotes intracellular Aβ42 accumulation co-localizing with TGN/endosomal markers provided early evidence that PS2-dependent cleavage preferentially occurs in intracellular acidic compartments.

   Evidence Stable SH-SY5Y lines with mutant APP/PS2, quantitative Western blot, immunofluorescence co-localization

   PMID:14725619

   Open questions at the time

   • The targeting mechanism sending PS2 to these compartments was unknown
   • Distinction from PS1-generated Aβ pools was not made
5. 2005 High

   Reconstitution of PS-null cells revealed that PS2—but not PS1—selectively restores PDGF receptor expression and PI3K/Akt–ERK signaling through its N-terminal fragment and the co-activator FHL2, establishing a γ-secretase-independent signaling function unique to PS2.

   Evidence PS-/- cell reconstitution with domain-deletion mutants, Co-IP of FHL2, Akt/ERK phosphorylation assays

   PMID:16014629

   Open questions at the time

   • Mechanism by which PS2-NTF/FHL2 controls PDGFR transcription was not elucidated
   • In vivo relevance in neural tissue was not tested
6. 2009 Medium

   Morpholino knockdown in zebrafish showed that Psen2 is required for Notch signaling during neurogenesis and neural crest development, and that Psen1 and Psen2 cooperate non-redundantly, clarifying PSEN2's developmental role.

   Evidence Zebrafish morpholino knockdown, in situ hybridization, Psen1/Psen2 double-knockdown epistasis

   PMID:19563801

   Open questions at the time

   • Morpholino off-target effects cannot be excluded without genetic mutants
   • Whether mammalian Psen2 has identical non-redundant developmental roles was not confirmed
7. 2016 High

   Identification of a unique PSEN2 cytoplasmic motif that recruits the AP-1 adaptor complex in a phosphorylation-dependent manner resolved why PS2-γ-secretase localizes to late endosomes/lysosomes, explaining the enrichment of longer Aβ species in these compartments and the amplified effect of FAD mutations.

   Evidence Subcellular fractionation, live-cell imaging, Co-IP/pulldown of AP-1, motif mutagenesis, Aβ isoform measurement

   PMID:27293189

   Open questions at the time

   • Kinase(s) phosphorylating the AP-1-binding motif were not identified
   • Whether therapeutic re-routing of PS2 away from endosomes would reduce Aβ42 was not tested
8. 2017 High

   CRISPR/Cas9 isogenic correction of PSEN2 N141I in iPSC-derived cholinergic neurons normalized both elevated Aβ42/40 and electrophysiological deficits, providing causal proof that this single mutation drives both biochemical and functional neuronal phenotypes in human cells.

   Evidence iPSC-derived BFCNs, ELISA, patch-clamp, CRISPR isogenic correction

   PMID:29078805

   Open questions at the time

   • Mechanism linking elevated Aβ42/40 to electrophysiological changes was not dissected
   • Long-term neurodegeneration phenotypes were not modeled
9. 2018 Medium

   Demonstrating that PSEN2 N141I disrupts calcium flux in iPSC-derived neurons—correctable by CRISPR and preventable by chronic insulin treatment—linked PSEN2 to ER calcium homeostasis and identified a potential therapeutic axis.

   Evidence iPSC-derived BFCNs, Ca²⁺ flux assays, CRISPR isogenic controls, chronic insulin treatment

   PMID:29945658

   Open questions at the time

   • Molecular target of insulin action on PSEN2-mediated Ca²⁺ leak was not identified
   • Whether WT PSEN2 also modulates ER Ca²⁺ at physiological levels was not resolved
10. 2019 High

    Mechanistic dissection showed that FAD-PSEN2 mutations impair autophagy by partially depleting ER Ca²⁺, reducing cytosolic Ca²⁺ transients, and preventing RAB7 recruitment to autophagosomes—a pathway independent of γ-secretase catalytic activity—connecting PSEN2's calcium role to a concrete downstream cellular process.

    Evidence Multiple FAD-PSEN2 cell models, autophagy flux assays, RAB7 imaging, FRET-based Ca²⁺ sensors, γ-secretase inhibitor controls, siRNA

    PMID:30892128

    Open questions at the time

    • How PSEN2 physically modulates ER Ca²⁺ leak channels remains unknown
    • Whether autophagy impairment contributes to neurodegeneration in vivo was not tested
11. 2020 Medium

    Expression of PSEN2 N141I in microglia impaired γ-secretase activity, exaggerated NF-κB-driven inflammatory responses, and altered microglial morphology in vivo, revealing a cell-type-specific contribution of mutant PSEN2 to neuroinflammation.

    Evidence Transgenic PSEN2 N141I mouse, primary microglia isolation, cytokine ELISA, NF-κB reporter, in vivo LPS challenge

    PMID:32741831

    Open questions at the time

    • Whether the inflammatory phenotype is γ-secretase-dependent or independent was not resolved
    • Contribution of microglial PSEN2 dysfunction to overall disease progression was not quantified
12. 2023 Medium

    Discovery of aberrant PSEN2 splicing events (cryptic exon, intron retention) specifically enriched in sporadic AD brain, generating truncated proteins and reducing canonical transcripts, suggested that PSEN2 loss-of-function extends beyond FAD mutations to sporadic disease.

    Evidence Long-read targeted isoform sequencing of prefrontal cortex from sporadic/familial AD and controls, replicated in independent cerebellum RNA-seq

    PMID:35949106

    Open questions at the time

    • Functional consequence of truncated PSEN2 splice products on γ-secretase and Ca²⁺ was not determined
    • Cause of aberrant splicing (e.g., RNA-binding protein dysregulation) was not identified

• Key unresolved questions include the identity of the kinase phosphorylating the PSEN2 AP-1-binding motif, the structural basis by which PSEN2 modulates ER Ca²⁺ independently of its catalytic site, and whether therapeutic targeting of PSEN2 compartment-specific γ-secretase activity or its Ca²⁺ function can alter disease trajectories in vivo.
• No kinase identified for the AP-1 motif phosphorylation
• Structural basis of γ-secretase-independent ER Ca²⁺ modulation unknown
• In vivo therapeutic proof-of-concept for compartment-selective PSEN2 targeting absent