🧬 Aging Wiki

dkk1

Properties1
aliasesDickkopf-1, Dkk-1, DKK-1, hDkk-1

10 min read

DKK1 (Dickkopf-1)

A small secreted Wnt antagonist that suppresses canonical Wnt signaling by triggering internalization of the co-receptor complex LRP5/6. DKK1 is elevated in several aging-relevant contexts — Alzheimer’s disease, chronic kidney disease, and multiple myeloma bone disease — making it a therapeutic target distinct from, but mechanistically parallel to, its sibling antagonist sclerostin. Anti-DKK1 antibodies (BHQ880, DKN-01) are in clinical development, though no aging-indication approval exists.

Identity

  • UniProt: O94907 (DKK1_HUMAN), Swiss-Prot reviewed
  • NCBI Gene: 22943
  • HGNC: HGNC:2891 (symbol DKK1)
  • Ensembl: ENSG00000107984
  • Chromosomal location: 10q21.1
  • Mouse ortholog: Dkk1 (one-to-one ortholog; highly conserved across vertebrates)
  • Length: 266 amino acids (canonical); signal peptide (aa 1–31) cleaved post-translationally
  • Molecular weight: ~29 kDa (mature form ~24 kDa after signal cleavage)
  • GenAge: not listed (DKK1 is not a curated GenAge aging gene; aging relevance is disease-context driven)

Domain structure

DKK1 contains two cysteine-rich domains (CRDs) stabilized by multiple disulfide bonds:

  • CRD1 (DKK-N domain, aa 85–138) — conserved in all DKK family members; weakly interacts with LRP6 E3/E4 domains but is not the primary binding site for LRP5/6 antagonism
  • CRD2 (Colipase-fold / C-terminal domain, aa 189–263) — the functional Wnt-inhibitory domain; binds the LRP5/6 E1/E2 propeller domains directly with high affinity; also the binding interface for Kremen1/2 co-receptors

The two CRDs are separated by a flexible linker. The C-terminal CRD2 alone is sufficient for Wnt pathway inhibition.

Mechanism — LRP5/6 antagonism via Kremen-mediated internalization

DKK1 inhibits canonical Wnt signaling through a stepwise internalization mechanism that is mechanistically distinct from sclerostin:

  1. DKK1 (secreted) binds the LRP5/6 E1/E2 extracellular propeller domains via CRD2 with nanomolar affinity
  2. Bound DKK1 recruits Kremen1 or Kremen2 (KREMEN1/2) transmembrane proteins to form a ternary DKK1–LRP5/6–Kremen complex
  3. The ternary complex is endocytosed and routed for lysosomal degradation, reducing cell-surface LRP5/6
  4. With LRP5/6 absent, Wnt-mediated Frizzled–LRP5/6 co-receptor assembly cannot occur → Axin–APC–GSK3β destruction complex remains active → β-catenin is phosphorylated, ubiquitinated, and degraded
  5. Net: suppressed TCF/LEF-mediated target gene transcription

Contrast with sclerostin (SOST): sclerostin binds LRP5/6 at the same E1/E2 propellers but via a different binding loop and does NOT recruit Kremen; its inhibition is steric/competitive and does not cause receptor internalization. DKK1-mediated LRP5/6 loss is therefore more durable per molecule than sclerostin-mediated steric blockade. See sost for the parallel comparison.

Tissue expression

Broadly expressed across adult tissues including bone (osteoblasts, osteocytes), brain (neurons, astrocytes), kidney, liver, and intestine. During embryogenesis DKK1 is a head-organizer signal (Xenopus anterior patterning). In adults the most aging-relevant expression is in:

  • Bone — osteoblast-derived DKK1 provides negative-feedback on Wnt-driven bone formation
  • Brain — neuronal DKK1 rises in Alzheimer’s disease brain and CSF
  • Kidney — DKK1 elevated in CKD, contributing to the bone-vascular dysregulation axis

Aging-relevant biology

Alzheimer’s disease

DKK1 is elevated in post-mortem brain samples and cerebrospinal fluid from AD patients relative to age-matched controls, and in transgenic AD animal models 1. In hippocampal neurons, Wnt signaling maintains dendritic spine density and synaptic protein clustering; DKK1 suppression of this signal causes synapse retraction. Key findings:

  • Amyloid-β increases the expression of DKK1 in neurons; DKK1 in turn causes rapid synaptic disassembly by blocking canonical Wnt/LRP6 signaling at synapses 1. Separately, Aβ also activates the non-canonical WNT5A-ROR2-JNK pathway, which independently reduces synapse formation; DKK1 acts as a parallel toxic stimulus that potentiates this effect — with DKK1 blockade and JNK blockade each partially rescuing synaptogenesis 2
  • DKK1-neutralizing antibodies protect hippocampal synapses from Aβ toxicity in rodent models 1
  • In SAMP8 accelerated-aging mice, DKK1 protein levels rise with age alongside reduced β-catenin activation and progressive neuronal loss 3

Preclinical anti-DKK1 antibody studies are promising but no human data exist. needs-human-replication

DimensionStatus
Pathway conserved in humans?yes — LRP5/6 and Kremen highly conserved
Phenotype (synaptic loss) conserved?partial — AD brain DKK1 elevation is human, but causal directionality unproven in humans
Replicated in humans?no — only observational; no anti-DKK1 trial in AD

Chronic kidney disease / mineral-bone disorder

DKK1 and sclerostin are both elevated in CKD, contributing to renal osteodystrophy and vascular calcification 4. In an in-vitro co-culture experiment, calcified aortic tissue from 5/6-nephrectomized Wistar rats secreted DKK1 (median 119 [78–551] pg/mL at 24 h; 353 [110–686] pg/mL at 48 h; not detected from normal aorta) and sclerostin into the culture media; co-incubation of this conditioned media with UMR-106 osteoblast-like cells completely abolished calcium crystal formation (Alizarin red staining; p<0.001 vs normal aorta co-culture), and upregulated the mineralization inhibitors osteopontin and ANKH 5. These findings support a direct vascular-to-bone crosstalk hypothesis in CKD-mineral and bone disorder (CKD-MBD), though the in-vivo causal contribution of DKK1 specifically (versus co-secreted sclerostin and activin A) remains to be disentangled. needs-human-replication for the directionality of the DKK1 contribution specifically.

Multiple myeloma bone disease

Multiple myeloma cells secrete high levels of DKK1, which suppresses osteoblast differentiation via Wnt inhibition and drives the lytic bone lesions characteristic of MM 6. The anti-DKK1 monoclonal antibody BHQ880 (Novartis) entered clinical trials for MM-related bone disease; phase 1/2 data showed increased bone formation markers. Trial activity for MM-bone indication appears to have stalled (no recent phase 3 completion); ongoing as of 2022–2023 literature. needs-current-trial-status

DKK1 circulating levels show age-associated increases in some cross-sectional studies. Unlike sclerostin (strongly osteocyte-restricted), DKK1 rises are less uniform and contextually driven (inflammatory disease, malignancy, CKD). A large RCT-quality longitudinal dataset on DKK1 trajectory in healthy aging does not yet exist. needs-human-replication

Therapeutic development (anti-DKK1 antibodies)

  • BHQ880 (Novartis; formerly MOR1404) — fully human anti-DKK1 IgG4; phase 1/2 evaluated in multiple myeloma bone disease. Demonstrated increases in P1NP (bone formation marker) and N-MID osteocalcin; clinical program status uncertain post-2022.
  • DKN-01 (Leap Therapeutics) — anti-DKK1 IgG4; currently 2 active (ACTIVE_NOT_RECRUITING) trials as of 2026-05-23 (ClinicalTrials.gov v2 API): NCT04166721 (phase 1/2, gastric/esophageal cancer, Royal Marsden NHS) and NCT05761951 (phase 2, endometrial cancer, MD Anderson); uses DKK1 as an immune-evasion target, not bone disease. Phase 1 trials in multiple myeloma (NCT01457417, NCT01711671) and biliary cancer (NCT02375880) are completed. No aging-indication program 7.

Aging-context druggability rationale (tier 2): Clinical-stage probes (BHQ880, DKN-01) exist with mechanistic evidence for aging-relevant indications (bone, brain), but no FDA-approved therapy exists for an aging indication that engages DKK1. Bone-specific romosozumab (anti-SOST) has a regulatory pathway that anti-DKK1 could follow for osteoporosis, but no phase 3 trial has completed.

Family context

The Dickkopf family includes DKK1, DKK2, DKK3, and DKK4. DKK1 and DKK2 both antagonize LRP5/6; DKK3 does not bind LRP5/6 directly (it modulates TGFR/Kremen interactions) — see the recent Science Advances study showing DKK3 suppresses Aβ clearance via LRP1, a distinct mechanism 8. DKK4 is a weak Wnt inhibitor. DKK1 is the most studied in aging contexts.

Pathway membership and interactors

  • Canonical pathway: wnt-beta-catenin — DKK1 is the primary secreted antagonist of LRP5/6 co-receptor activity
  • Co-receptors: Kremen1/Kremen2 (KREMEN1/2) — required for DKK1-mediated LRP5/6 internalization
  • Mechanistic parallel: sost — shares the LRP5/6 target but uses steric rather than internalization mechanism
  • Upstream interaction: klotho — some evidence that FGF23/Klotho signaling modulates DKK1 in the CKD context; contested; see klotho for evidence
  • Downstream effects on bone: bone (remodeling suppression); osteoporosis (via Wnt inhibition-driven osteoblast suppression)
  • Downstream effects in brain: alzheimers-disease (synaptic Wnt loss contribution)
  • Inflammation context: chronic-inflammation — inflammatory cytokines (TNFα, IL-6) can upregulate DKK1; DKK1 suppression of Wnt in immune cells also affects macrophage polarization (bidirectional)

Gaps and limitations

  • needs-human-replication — anti-DKK1 strategy in AD has no human trial data; mouse models have poor translational record for synaptic rescue (see alzheimers-disease for extrapolation caveats)
  • needs-human-replication — DKK1’s causal role in CKD-driven vascular calcification is mechanistically plausible but not established by RCT or MR in humans
  • needs-current-trial-status — BHQ880 clinical program status post-2022 is unclear; no completed phase 3
  • dose-response-unclear — circulating DKK1 levels in aging: no established reference range for “aging-relevant” elevation vs noise
  • [[osteoblasts]] page exists (seeded 2026-06-02; unverified); DKK1’s most mechanistically documented target cell
  • [[lrp5-lrp6]] page exists (seeded 2026-05-23; unverified)
  • [[sost]] page exists (seeded 2026-05-23; unverified)

Footnotes

Footnotes

  1. doi:10.1093/jmcb/mjt049 · Purro SA, Galli S, Salinas PC · J Mol Cell Biol 2014 · review (not a primary research paper) · Reviews the role of Wnt signaling in synaptic maintenance and AD. Primary data for DKK1 elevation in AD brain and DKK1-neutralizing antibody synapse rescue comes from: Purro SA et al. J Neurosci 2012, 32:3492–3498 (rat hippocampal neuron culture + mouse brain slice model); and Caricasole et al. J Neurosci 2004, 24:6021–6027 (human AD post-mortem brain samples + transgenic mouse models). The review reports that Aβ increases DKK1 expression and that anti-DKK1 antibody blockade suppresses Aβ-mediated synaptotoxicity. · archive: downloaded (PMC4344549) ↩ ↩2 ↩3

  2. doi:10.1242/jcs.263526 · Fang K et al. · J Cell Sci 2025 · in-vitro (SH-SY5Y human neuroblastoma + iPSC-derived cortical neurons with PSEN1^A75V mutation) · The primary pathway studied is WNT5A-ROR2→JNK (non-canonical Wnt/PCP), which reduces synapse formation; Aβ₁₋₄₂ oligomers and DKK1 treatment each upregulate ROR2 (~1.5-fold and ~1.53-fold mRNA respectively) and activate JNK (~3.02- and ~3.04-fold C:N reporter shift) — DKK1 and Aβ are parallel toxic stimuli converging on ROR2-JNK; DKK1 is not the intermediary step in the Aβ→JNK pathway. Blocking ROR2 or JNK partially rescues synaptogenesis from both Aβ and DKK1 treatment. · archive: downloaded ↩

  3. doi:10.1016/j.neurobiolaging.2014.09.017 · Bayod S et al. · Neurobiol Aging 2015 · in-vivo (SAMP8 accelerated-aging mouse) · DKK1 protein rises with age in SAMP8 hippocampus alongside reduced β-catenin and neuronal loss · archive: not OA no-fulltext-access — claims not independently verified against full PDF ↩

  4. doi:10.1007/s00223-020-00803-2 · Carrillo-López N et al. · Calcif Tissue Int 2021 · review · DKK1 and SOST are elevated in CKD; role in bone-vascular Wnt axis dysregulation · archive: not OA no-fulltext-access — abstract-level only; claims not independently verified ↩

  5. doi:10.1002/jbm4.10610 · Mace ML et al. · JBMR Plus 2022 · in-vitro vascular-bone co-culture · model: 5/6-nephrectomized male Wistar rats (22 weeks old; n=6 uremic, n=6 normal) aorta rings co-cultured with UMR-106 osteoblast-like cells · DKK1 secretion: not detected in normal aorta; uremic calcified aorta secreted median 119 [78–551] pg/mL at 24h and 353 [110–686] pg/mL at 48h (p<0.001 vs normal aorta at 48h, Mann-Whitney U). Uremic aorta co-incubation completely abolished UMR-106 calcium crystal formation (Alizarin red; p<0.001 vs normal aorta co-culture; experiment repeated 3× with n=24 total wells per group). Also upregulated UMR-106 osteopontin (Spp1 mRNA ~25.5-fold vs control; p<0.001) and ANKH mRNA (~2.95-fold; p<0.005). Sclerostin and activin A were co-secreted; the relative contribution of DKK1 vs these co-factors not individually disambiguated. LiCl (Wnt/β-catenin activator) partially rescued osteopontin upregulation, implicating Wnt pathway suppression in the mineralization inhibition. · archive: downloaded (PMC9009125) ↩

  6. doi:10.3390/ijms23063112 · Gau YC et al. · Int J Mol Sci 2022 · review · MM cells secrete DKK1 (stimulated by IL-6 from the bone marrow microenvironment) → suppresses osteoblast Wnt/β-catenin activity → lytic bone disease; DKK1 is one of multiple osteolytic factors (alongside RANKL, sclerostin, activin A, CCL3, CCL20, annexin II, IL-3, IL-6, IL-17, TNF, BAFF). BHQ880 phase 1/2 MM bone-disease trial summarized. DKN-01 not discussed in this review. · archive: downloaded ↩

  7. doi:10.3390/ijms27093780 · Canella A et al. · Int J Mol Sci 2026 · review · DKK1 in tumor immune evasion; DKN-01 (anti-DKK1) oncology pipeline summary · archive: not confirmed in archive (DOI not found) ↩

  8. doi:10.1126/sciadv.adz2099 · Yang R et al. · Sci Adv 2025 · in-vitro + in-vivo (transgenic AD mice) · DKK3 (not DKK1) inhibits Aβ clearance via LRP1; compound SJ-300 rescues cognition — distinct from DKK1-LRP5/6 mechanism · archive: pending download ↩

Graph View

Backlinks