ACAN

ACAN encodes aggrecan, the principal chondroitin sulfate proteoglycan of cartilage extracellular matrix and a core structural component of perineuronal nets in the brain. In cartilage, ACAN transcription is driven by SOX9 binding to multiple redundant enhancers distributed across >100 kb, with SOX9 occupancy gated by SIRT1-mediated deacetylation that promotes nuclear import and by TET1-dependent 5-hydroxymethylcytosine deposition at SOX9 binding sites; additional transcriptional input comes from SHOX/SHOX2 cooperating with SOX5/SOX6 and from TGF-β1 signaling through the mTOR/4E-BP1–Smad axis (PMID:22820679, PMID:26910618, PMID:33134768, PMID:24421874, PMID:32485037). Loss-of-function mutations—including frameshifts causing nonsense-mediated decay and G3-domain missense variants that impair secretion and ECM ligand binding—cause a spectrum of aggrecanopathies encompassing spondyloepiphyseal dysplasia, familial short stature, and osteochondritis dissecans through haploinsufficiency (PMID:16080123, PMID:17952705, PMID:35338222, PMID:38782218). In the brain, aggrecan integrates into perineuronal nets via its G1 domain hyaluronan-binding groove and exerts cell-type-specific functions: it provides neuroprotection against pathological tau in neurons, and conditional deletion from PV+ interneurons or CA2 pyramidal neurons selectively impairs fear memory, social memory, and reversal learning (PMID:32737917, PMID:37131076).

1. 2005 High

   Identifying a frameshift mutation in ACAN cosegregating with spondyloepiphyseal dysplasia type Kimberley established ACAN as a human skeletal disease gene, resolving the question of whether aggrecan deficiency could cause Mendelian bone disease.

   Evidence Mutation screening and cosegregation analysis in a multigenerational SEDK family mapped to 15q26.1

   PMID:16080123

   Open questions at the time

   • Mechanism of disease limited to genetic linkage; no in vitro functional assay of the truncated protein
   • Whether haploinsufficiency or dominant-negative effects drive pathology was unresolved
2. 2007 High

   Demonstrating that a frameshift allele in Dexter cattle undergoes nonsense-mediated decay (retaining only 8% mutant mRNA) established haploinsufficiency as the primary disease mechanism for ACAN loss-of-function alleles.

   Evidence mRNA quantification in heterozygous chondrocytes from Dexter cattle with a 4-bp ACAN insertion; genotyping in worldwide pedigrees

   PMID:17952705

   Open questions at the time

   • Whether human SEDK mutations similarly undergo NMD was not tested
   • Protein-level consequences of heterozygous loss not quantified
3. 2012 High

   Functional testing of conserved non-coding sequences revealed that ACAN transcription in cartilage is controlled by at least eleven redundant enhancers spanning >100 kb, many harboring SOX9 sites, explaining the robustness of cartilage-specific expression.

   Evidence Transgenic zebrafish enhancer reporter assays for 24 conserved non-coding sequences around the ACAN locus

   PMID:22820679

   Open questions at the time

   • Individual enhancer necessity (deletion studies) not performed
   • Combinatorial or hierarchical relationships among enhancers unknown
4. 2014 High

   Showing that SHOX2 activates ACAN transcription through physical interaction with SOX5/SOX6 established a cooperative SOX trio–SHOX axis in growth plate chondrocytes, explaining how short stature transcription factors converge on the ACAN promoter.

   Evidence Luciferase reporter, yeast-two-hybrid, co-immunoprecipitation, and immunohistochemistry of human fetal growth plates

   PMID:24421874

   Open questions at the time

   • Whether SHOX2 contacts specific ACAN enhancers (ChIP) was not tested
   • Relative contribution of SHOX vs SHOX2 to endogenous ACAN expression unclear
5. 2016 High

   Revealing that SIRT1-mediated deacetylation of SOX9 promotes its nuclear import via importin-β and increases SOX9 binding to the ACAN -10 kb enhancer explained how metabolic (NAD+) status gates chondrogenic gene expression.

   Evidence Co-IP of SOX9–SIRT1, ChIP at ACAN enhancer, pharmacological perturbation with NAD/SIRT1 inhibitors, and importazole in primary human OA chondrocytes

   PMID:26910618

   Open questions at the time

   • Whether SIRT1 deacetylation acts at acetylation sites relevant in vivo remains to be mapped
   • Contribution of other HDACs to SOX9 activity at ACAN not addressed
6. 2017 High

   A mouse knock-in model with ~50% reduction in Acan expression recapitulated dwarfism with shorter bones, reduced growth plates, and thinner articular cartilage, formally demonstrating that Acan gene dosage is critical for skeletal growth.

   Evidence Homozygous Agc1CreERT/CreERT knock-in mice with histology, proteoglycan staining, and gene expression analysis

   PMID:28921880

   Open questions at the time

   • Heterozygous phenotype not fully characterized
   • Molecular compensation by other proteoglycans not examined
7. 2020 High

   Three studies collectively deepened understanding of ACAN regulation and neural function: TET1-dependent 5hmC deposition was shown to be required for SOX9 occupancy at the Acan locus; the mTOR/4E-BP1 pathway was found to control ACAN transcription via translational regulation of inhibitory Smads; and in a tauopathy mouse model, aggrecan in perineuronal nets was shown to shield neurons from pathological tau internalization.

   Evidence Tet1 knockdown with ChIP and 5hmC mapping in ATDC5 cells; mTOR/4E-BP1 silencing with Smad ChIP in OA chondrocytes; bigenic TauP301L×Acan mouse with IHC, Western blot, and ELISA

   PMID:32485037 PMID:32737917 PMID:33134768

   Open questions at the time

   • Whether TET1 acts directly at the ACAN locus or through broader chromatin remodeling is unclear
   • The mTOR–Smad–ACAN axis lacks in vivo validation
   • The neuroprotective mechanism (physical shielding vs. signaling) is not resolved
8. 2022 High

   Functional studies of G3-domain missense variants linked to familial osteochondritis dissecans demonstrated that G3 integrity is required for aggrecan secretion and ECM ligand binding, establishing a distinct pathogenic mechanism from haploinsufficiency.

   Evidence Recombinant variant protein secretion assays, ECM ligand binding assays, and analysis of full-length aggrecan from heterozygous patient cartilage

   PMID:35338222

   Open questions at the time

   • Structural basis for secretion defect not determined
   • Whether retained protein triggers ER stress was not examined
9. 2023 High

   Conditional Hdac2 deletion from PV+ interneurons reduced Acan expression and perineuronal net density, and direct siRNA knockdown of Acan recapitulated impaired fear memory recovery, establishing Acan as a functional effector of HDAC2-regulated PV+ cell maturation and fear extinction.

   Evidence PV+-specific Hdac2 conditional KO, pharmacological HDAC inhibition, intravenous Acan siRNA, immunofluorescence for PNNs, and behavioral fear conditioning in mice

   PMID:37131076

   Open questions at the time

   • Which HDAC2-regulated transcription factor directly controls Acan in PV+ cells is unknown
   • Whether HDAC2–Acan axis operates outside amygdala/PFC circuits not tested
10. 2024 High

    A non-canonical intronic ACAN splice variant was functionally validated by minigene assay, expanding the mutational spectrum of aggrecanopathies to include deep-intronic variants causing familial short stature.

    Evidence Sanger sequencing in a pedigree with short stature; minigene splicing assay demonstrating cryptic splice site activation and frameshift

    PMID:38782218

    Open questions at the time

    • Endogenous mRNA from patient cells not characterized
    • Whether NMD degrades the aberrant transcript was not confirmed

• Key unresolved questions include the precise structural basis for how G3-domain variants impair aggrecan secretion, the identity of the transcription factor directly linking HDAC2 to Acan in PV+ neurons, and whether HA-binding-independent mechanisms contribute substantially to aggrecan integration into perineuronal nets in vivo.
• No full-length aggrecan structure available
• Cell-type-specific transcriptional regulation in neural contexts poorly defined
• Relative contribution of HA-dependent vs HA-independent PNN integration not quantified in vivo