MMP24

MMP24 (MT5-MMP) is a brain-enriched membrane-type matrix metalloproteinase that couples extracellular and intracellular proteolytic functions to neural stem cell regulation, nociception, and amyloidogenic APP processing (PMID:19805319, PMID:24952463, PMID:26202697). Its catalytic zinc-binding domain activates progelatinase A (MMP2) and preferentially degrades proteoglycans among extracellular matrix components, an activity abolished by mutation of the HEXGH zinc-binding motif and sensitive to TIMP-2 (PMID:10085137, PMID:10363975, PMID:10622708). Surface presentation of the enzyme is controlled by recycling: its C-terminal EWV motif binds the PDZ adaptor Mint-3 to retrieve internalized MT5-MMP back to the plasma membrane, and loss of this signal or Mint-3 depletion reduces surface activity (PMID:14990567). Activation requires PCSK6, which recognizes and cleaves the RRRNKR propeptide motif to generate active enzyme (PMID:38216110). In neurons, MT5-MMP sheds the N-cadherin ectodomain to regulate adult neural stem cell activation and quiescence and to modulate nociceptor neuro-immune interactions, with knockout mice showing altered thermal sensitivity and impaired stem cell activation (PMID:19805319, PMID:24952463). Acting as an η-secretase, MT5-MMP associates with APP and promotes amyloidogenic processing, raising Aβ, C99, and sAPP95 without altering α-, β-, or γ-secretase activities, in part by directing APP and C99 into the endolysosomal system (PMID:26202697, PMID:28119565, PMID:34117802). This pro-amyloidogenic activity is only partly proteolytic: the C-terminal non-catalytic and intracellular domains govern C99 sorting and stability through brain-derived binding partners including N4BP2L1, EIG121, BIN1, and TMX3, and deletion of these domains routes C99 to proteasomal degradation [PMID:34117802, PMID:38345408, PMID:bio_10.1101_2025.05.22.655451].

1. 1999 High

   Established MT5-MMP as a catalytically active metalloproteinase capable of activating a downstream proteinase, defining its core enzymatic function and zinc-dependent mechanism.

   Evidence Co-expression in MDCK cells with active-site (Glu→Ala) mutagenesis, plus in vitro assays with recombinant catalytic domain and purified soluble enzyme

   PMID:10085137 PMID:10363975 PMID:10622708

   Open questions at the time

   • Physiological MMP2 activation in vivo not demonstrated
   • Full ECM substrate repertoire beyond proteoglycans not defined
2. 1999 Medium

   Localized MT5-MMP to the plasma membrane and showed it can shed as a soluble proteinase, distinguishing it from membrane-retained MT-MMPs.

   Evidence Immunofluorescence and Western blot of transfected COS-7 and conditioned-media analysis

   PMID:10085137 PMID:10363975

   Open questions at the time

   • Shedding mechanism and protease responsible not identified
   • Relative balance of membrane vs soluble pools in neurons unknown
3. 2004 High

   Identified the trafficking logic controlling surface activity, showing a C-terminal PDZ-binding motif recruits Mint-3 to recycle the enzyme to the membrane.

   Evidence Yeast two-hybrid, EWV deletion mutagenesis, Mint-3 siRNA, surface activity assays

   PMID:14990567

   Open questions at the time

   • Endocytic route and other adaptors not mapped
   • Link to APP processing not yet examined
4. 2009 High

   Connected MT5-MMP proteolysis to physiology by showing N-cadherin cleavage in nociceptors controls neuro-immune interactions and pain sensitivity.

   Evidence Mmp24 knockout mice, behavioral nociception assays, N-cadherin cleavage analysis

   PMID:19805319

   Open questions at the time

   • Direct vs indirect N-cadherin cleavage in vivo not fully separated
   • Mast-cell interaction mechanism incompletely defined
5. 2014 High

   Extended the N-cadherin-shedding function to adult neural stem cells, establishing a role in stem cell activation and quiescence.

   Evidence MT5-MMP knockout mice, N-cadherin shedding and NSC functional assays in vivo

   PMID:24952463

   Open questions at the time

   • Downstream signaling from N-cadherin shedding not resolved
   • Dispensable for NSC identity—upstream regulators unknown
6. 2015 High

   Defined MT5-MMP as an η-secretase that physically associates with APP and drives amyloidogenic processing independent of classical secretases.

   Evidence Reciprocal Co-IP and co-localization in HEKswe cells, MT5-MMP KO × 5xFAD cross, ELISA/Western

   PMID:26202697

   Open questions at the time

   • Precise cleavage site generating sAPP95 not mapped here
   • Causal contribution to disease pathology not established
7. 2017 Medium

   Linked the pro-amyloidogenic effect to subcellular trafficking by showing MT5-MMP localizes to early endosomes and routes APP/C99 there.

   Evidence Immunofluorescence, subcellular fractionation, Aβ40 ELISA in HEKswe cells

   PMID:28119565

   Open questions at the time

   • Single cell model
   • Trafficking machinery directing endosomal sorting unidentified
8. 2021 High

   Dissected proteolytic from non-proteolytic contributions, showing the C-terminal domains sort C99 to the endolysosomal system and protect it from proteasomal degradation even without catalytic activity.

   Evidence Domain deletion and catalytic-dead mutagenesis, fractionation, proteasome inhibitors, Western blot in HEKswe cells

   PMID:34117802

   Open questions at the time

   • Direct C99-binding interface not structurally defined
   • Single cell system
9. 2021 Medium

   Revealed transcriptional/post-transcriptional control in pain, with FTO-mediated m6A demethylation enhancing MMP24 translation to drive ERK-dependent hypersensitivity.

   Evidence MeRIP-seq, RIP, FTO overexpression/knockdown, ERK assays, neuropathic pain behavior

   PMID:33995105

   Open questions at the time

   • Direct enzymatic substrate driving ERK activation unidentified
   • Single lab
10. 2024 High

    Identified the activating protease, showing PCSK6 cleaves the RRRNKR propeptide motif to generate active MT5-MMP and enable amyloidogenic processing.

    Evidence Co-IP, cleavage-motif mutagenesis, PCSK6 knockdown in N2AAPP cells and AAV KD in APP23/PS45 mice

    PMID:38216110

    Open questions at the time

    • Whether PCSK6 is the sole physiological activator unknown
    • Spatial site of activation not defined
11. 2024 Medium

    Mapped organelle-dependence of η-secretase activity and identified C-terminal-domain binding partners that enhance sAPPη production.

    Evidence Subcellular-targeted fusion constructs with sAPPη ELISA; yeast two-hybrid screen of brain cDNA with functional follow-up

    PMID:38345408

    Open questions at the time

    • Direct vs indirect nature of N4BP2L1/EIG121/BIN1/TMX3 effects not resolved
    • Endogenous relevance of nuclear/cytosolic localization unclear
12. 2025 Medium

    Resolved the intracellular domain as a trafficking determinant, showing specific IC mutations divergently control C99 sorting, degradation, and Aβ output.

    Evidence IC-domain mutagenesis, high-content imaging, proximity ligation assay in human cell lines (preprint)

    PMID:bio_10.1101_2025.05.22.655451

    Open questions at the time

    • Preprint, not yet peer-reviewed
    • Specific motifs/adaptors mediating each IC-cluster effect not defined

• How MT5-MMP's dual proteolytic and trafficking-scaffold functions are coordinated in vivo and whether targeting them alters Alzheimer pathology remains unresolved.
• No structural model of the C-terminal/IC domain interactions with C99
• Therapeutic modulation untested
• Integration of pain, NSC, and amyloid roles in the same neurons unknown