CD44

CD44 is a multifunctional type-1 transmembrane glycoprotein that serves as the principal cell-surface receptor for hyaluronic acid (HA). A single human gene at chromosome 11p13 encodes up to 19 alternative splice isoforms: the standard form (CD44s, ~85 kDa core) and variant isoforms (CD44v3–v10) that carry additional exon sequences in the extracellular stem region. CD44 mediates HA-dependent cell adhesion, mechanosensing, and signal transduction — most critically the contact-inhibition axis via NF2 (merlin) → Hippo/YAP — and participates in immune-cell activation (T-cell costimulation, macrophage polarization), hematopoietic stem cell homing, and stem cell niche interactions. It is the receptor through which the naked mole-rat HMW-HA cancer-resistance mechanism operates (Tian 2013) ¹, and it carries a pan-species longevity correlation independent of body mass that was quantified by Takasugi et al. 2023 ². CD44v-isoform expression is also the canonical cancer stem cell marker in multiple epithelial tumor types.

Identity

Field	Value
UniProt	P16070 (CD44_HUMAN; Swiss-Prot/reviewed)
NCBI Gene	960
HGNC	1681
Ensembl	ENSG00000026508
Chromosome	11p13
Canonical protein length	742 aa (isoform 1 / standard)
Mouse ortholog	Cd44 · UniProt P15379 · Ensembl ENSMUSG00000005087
Blood group system	Indian (IN) blood group antigen — CD44 is the carrier protein
GenAge ID	null (not in GenAge human aging gene set)

Isoforms. CD44s (standard; lacks variable exons) is the ubiquitous form. CD44v isoforms insert one or more of exons v3–v10 into the membrane-proximal stem stalk. CD44v6 and CD44v8–10 are the most clinically studied variants (cancer stem cell biology). Most cell-biology and aging studies use pan-CD44 antibody reagents and do not distinguish isoforms — a persistent analytical gap ³. needs-isoform-resolution

Structural domains

Domain	Residues (UniProt P16070)	Function
Signal peptide	1–20	Targeting to secretory pathway
HA-binding link domain	~32–120	HA binding; conserved Link module fold; weak affinity in monomer
Extracellular stem/stalk	~120–649	Variable exon insertion site (v3–v10); glycosylation; cleavage sites for MMPs/ADAM10
Transmembrane helix	650–670	Single-pass
Cytoplasmic tail	671–742	ERM (ezrin/radixin/moesin) binding; actin cytoskeleton linkage; PKC phosphorylation sites (Ser-672, Ser-686, Ser-706); substrate for Src kinase

Key PTMs (UniProt P16070): N-linked glycosylation at Asn-25, -57, -100, -110, -120, -350, -548, -599, -636; O-linked chondroitin sulfate at Ser-180; PKC-dependent phosphorylation at Ser-672 (UniProt P16070; Ponta 2003 discusses PKC-dependent serine phosphorylation of the cytoplasmic tail but uses variant-isoform residue numbering — the UniProt canonical isoform 1 positions are authoritative) ³. Extensive glycosylation expands the apparent molecular weight on gel from ~85 kDa (core) to 80–200 kDa depending on tissue and isoform.

HA-binding biology — MW-dependent receptor clustering

Monomeric CD44-HA binding is weak (Kd ~µM range) and is insufficient to activate downstream signaling. Functional signaling requires cross-linking of multiple CD44 molecules by long polymeric HA chains (HMW-HA, ≥500 kDa), which assembles receptor clusters in the plane of the membrane ³.

HMW-HA → CD44 clusters → NF2/merlin → contact inhibition

This is the central mechanism of naked mole-rat cancer resistance (Tian 2013) ¹:

1. HMW-HA (NMR cells produce HA chains 6–12 MDa, vs 0.5–3 MDa in mouse/guinea pig and 0.5–2 MDa in human fibroblasts) cross-links CD44 into stable surface clusters.
2. Clustered CD44 activates NF2 (merlin) — the neurofibromatosis type 2 tumor suppressor. Tian 2013 describes NF2 as mediating contact inhibition downstream of CD44; the specific Hippo kinase cascade (STK4/STK3 → LATS1/2 → YAP/TAZ) is a known downstream effector of NF2 in the broader literature but is not explicitly named in Tian 2013 ¹.
3. Active NF2/merlin promotes early contact-inhibition of proliferation.
4. Contact inhibition in NMR cells triggers p16^INK4a induction — a distinctively early growth arrest relative to human or mouse cells.
5. CD44 blockade in NMR cells (via a CD44-blocking antibody) abrogated the early contact-inhibition phenotype, confirming CD44 as a required node rather than a correlative marker ¹.

Dimension	Status	Notes
Pathway conserved in humans?	yes	Human CD44, NF2, and Hippo pathway are fully conserved; same signaling architecture
Phenotype conserved in humans?	partial	HMW-HA is less abundant in human tissues; full NMR-level contact-inhibition phenotype not observed
Replicated in humans?	no	No human intervention test; Zhang 2023 mouse nmrHas2 is the closest proxy 4

LMW-HA → pro-inflammatory CD44/TLR signaling

Low-molecular-weight HA fragments (produced during ECM turnover, tissue injury, and chronic inflammation) form fewer CD44 clusters. They instead act as damage-associated molecular patterns (DAMPs) that signal through TLR4 ⁵. Termeer 2002 showed that DCs from TLR-2-deficient mice were still susceptible to LMW-HA-induced maturation, demonstrating that TLR2 is not required; TLR4 is the primary receptor:

• LMW-HA → TLR4 → p38/p42/44 MAPK and NF-κB translocation → TNF-α production and DC maturation
• This LMW/TLR4 axis is mechanistically distinct from the HMW/CD44-cluster axis
• In aging, chronic tissue injury and elevated hyaluronidase activity progressively shift the HA molecular weight distribution toward LMW — amplifying TLR4/NF-κB signaling and attenuating the contact-inhibitory HMW-CD44/NF2/Hippo signal ⁵

The MW-dependent outcome (HMW → anti-proliferative / anti-inflammatory; LMW → pro-inflammatory) is the molecular basis for HA’s molecular-weight-dependent biology. See hyaluronic-acid for the broader MW-dependent biology discussion.

Aging context

Tian 2013: NMR HMW-HA → CD44 → NF2 → cancer resistance

Tian et al. 2013 ¹ demonstrated that:

• NMR cells synthesize HA chains of 6–12 MDa (vs 0.5–3 MDa in mouse/guinea pig and 0.5–2 MDa in human), primarily due to elevated Has2 expression and lower hyaluronidase activity
• NMR cells exhibit early contact inhibition at lower cell density than mouse or human cells
• CD44 blockade (via a CD44-blocking antibody) abolished this early contact-inhibition advantage, placing CD44 causally in the cancer-resistance pathway; NMR cells grown with CD44 antibodies reached higher cell density before arresting ¹
• The sequence NMR-HMW-HA → CD44 → NF2 (merlin) → p16^INK4a constitutes the principal molecular explanation for NMR cancer resistance

Zhang 2023: nmrHas2 mouse — CD44 as effector node

Zhang et al. 2023 ⁴ introduced the NMR Has2 gene into mice under a CAG-loxP-STOP-loxP inducible system (nmrHas2 transgenic, C57BL/6 background). In aged nmrHas2 mice, HA abundance and molecular weight were elevated in multiple tissues, producing downstream phenotypes consistent with CD44-mediated signaling:

• +4.4% median lifespan, +12.2% maximum lifespan; spontaneous cancer mortality reduced (57% vs 70% in controls; 49% vs 83% in survivors past 27 months)
• Attenuated inflammaging: reduced plasma IL-12p40, MIP1α, MIP1β, CCL7 in aged females; macrophages shifted toward M2-like polarization (higher Arg1/IL-10, lower IL-1β/IL-6 after LPS)
• Preserved gut barrier integrity
• Transcriptomic profile enriched for anti-senescence and longevity-associated pathways

Important caveat: Zhang 2023 does not directly dissect the CD44 mechanism in the mouse model. The paper focuses on whole-animal phenotypes and inflammatory readouts. The CD44-NF2 axis is the established upstream mechanism from Tian 2013, but Zhang 2023 does not provide independent CD44 knockdown or NF2 manipulation in the mouse model. Wikilink: zhang-2023-nmrhas2-mouse-healthspan.

Takasugi 2023: CD44 longevity correlation and ATF6 ER-stress resistance

Takasugi et al. 2023 (Cell Reports) ² isolated oligodendrocyte progenitor cells (OPCs) from NMRs and compared their transcriptome with those of short- and long-lived mammals. CD44 expression correlated positively with species maximum lifespan across fibroblasts of 10 mammalian species, liver of 27 mammalian species, and skin of 20 rodent species (phylogenetic generalized least squares, p < 0.0064 in fibroblasts; p < 0.05 in 85/100 liver tests). Mechanistically, CD44 localizes to the endoplasmic reticulum (ER) and enhances basal ATF6 (activating transcription factor 6) activity and ER stress resistance — independently of HA (the effect was observed in the absence of exogenous HA). CD44 siRNA knockdown reduced survival under tunicamycin-induced ER stress in both NMR and mouse OPCs; CD44 overexpression enhanced tunicamycin resistance. This HA-independent role of CD44 in proteostasis regulation adds a second downstream aging-relevant arm beyond the NF2 contact-inhibition axis. Note that CD44 was among 13 ECM-related genes elevated in both NMR and human OPCs relative to mouse/rat; the longevity correlation was driven by CD44 expression level, not HA abundance in the tissue ².

Aging context — immune compartment

CD44 is expressed on all hematopoietic lineages and is a central node in immune-cell aging ³:

T cells. CD44^hi expression distinguishes effector and memory T cells from naive. During thymic involution and peripheral T-cell aging, the CD44^hi:CD44^lo ratio skews progressively toward CD44^hi (antigen-experienced / effector-memory bias), shrinking the naive compartment and contributing to immunosenescence. CD44 itself is a costimulatory molecule that lowers the T-cell activation threshold.

Macrophages. HMW-HA → CD44 signaling in macrophages promotes M2-like polarization (anti-inflammatory; arginase-1^hi, IL-10^hi). Zhang 2023 bone-marrow-derived macrophage data directly demonstrated this shift in the nmrHas2 model ⁴. Conversely, LMW-HA → TLR4 in macrophages drives M1-like activation — a feedforward mechanism by which aging tissue ECM (increasingly LMW due to elevated hyaluronidase activity) amplifies inflammaging.

HSC homing. CD44 interacts with E-selectin on bone marrow sinusoidal endothelium (especially the HCELL glycoform — hematopoietic cell E-/L-selectin ligand — which is CD44 decorated with sialyl-Lewis^X). This CD44/E-selectin axis is required for HSC homing and engraftment. Age-related changes in this glycoform pattern may contribute to altered HSC trafficking seen in aged marrow. See stem-cell-exhaustion.

Microglia. CD44 is upregulated on reactive (disease-associated) microglia in neurodegeneration and aging brain. CD44 marks the microglial DAM (disease-associated macrophage) and homeostasis-exit states. Whether this upregulation is causal or a consequence of neuroinflammation is unresolved. no-mechanism

Cancer biology and the CD44 tractability problem

CD44 variant isoforms — particularly CD44v6 and CD44v8–10 — mark cancer stem cells (CSCs) in multiple epithelial tumor types. Dalerba et al. 2007 ⁶ established EpCAM^hi/CD44^+ as the tumor-initiating phenotype in human colorectal cancer. Using 6 distinct CRC xenograft lines established from patient specimens, EpCAM^hi/CD44^+ cells consistently formed tumors in NOD/SCID mice (200–500 cells sufficient), while EpCAM^low/CD44^− fractions consistently failed to engraft (0/5 injections at 10,000 cells). The co-marker is EpCAM^hi alongside CD44^+; CD44 alone is not sufficient as a CSC identifier in this study. CD44v6 is the principal isoform expressed on gastric and head/neck CSC populations.

CD44 has been one of the most-attempted but consistently-failing oncology targets:

• Bivatuzumab (anti-CD44v6 antibody-drug conjugate; MMAE payload) — Phase 1 in head/neck and breast cancer showed dose-limiting skin toxicity from on-target CD44 expression in normal stratified epithelium. Program terminated.
• RG7356 / RO5429083 (anti-CD44s humanized IgG1; Roche) — Phase 1 in advanced solid tumors ⁷ (65 patients, June 2011–November 2013). 21% stable disease (13/61 evaluable; median 12 weeks, range 6–35 weeks); no partial or complete responses. Dose-limiting toxicities: grade 3 headache (1,500 mg q2w; 1,350 mg qw) and grade 4 febrile neutropenia (2,250 mg q2w). MTD for q2w: 1,500 mg; MTD for qw not defined (study terminated early). No significant clinical benefit; program terminated for this indication.

Implications for aging: The consistent Phase 1 failures of CD44-directed antibodies in oncology indicate that direct CD44 antagonism at the extracellular domain is not clinically tractable as currently approached. For aging biology, the mechanistically-interesting direction is HA-MW augmentation acting through CD44 (promoting HMW-HA signaling) — not CD44 protein targeting. This is the approach of the Zhang 2023 / nmrHas2 paradigm, which targets the upstream HA-synthesis step rather than the receptor itself. contradictory-evidence on whether CD44 is a tractable direct therapeutic target.

Cellular senescence

CD44 is upregulated on certain senescent cell populations (senescent fibroblasts, senescent vascular smooth muscle cells), in part as a consequence of SASP-secreted cytokines that induce CD44 expression in a paracrine manner. The relationship is bidirectional:

• HMW-HA → CD44 → NF2/Hippo → p16^INK4a: can drive senescence entry in rapidly proliferating cells experiencing contact pressure; in the NMR context this is protective (early arrest before malignant transformation), but in the context of chronic HMW-HA in aged tissue may contribute to senescence accumulation in normal proliferating compartments. This is a mechanistic ambiguity not yet resolved. no-mechanism
• LMW-HA → CD44/TLR4 → NF-κB → SASP amplification: LMW-HA fragments generated during chronic inflammation signal through CD44 and TLR4 to amplify NF-κB, which is a master transcriptional driver of SASP. This creates a feedforward loop: senescent cells secrete matrix metalloproteinases → degrade HMW-HA → generate LMW-HA → TLR4/NF-κB → more SASP.

The net effect of CD44 on senescence therefore depends critically on which HA species is present — a reason to track HA molecular weight distribution in aged tissues, not just total HA abundance.

Druggability (Open Targets, 2026-05-12)

Open Targets Platform (ENSG00000026508):

• Small molecule (SM): All tractability buckets false (no SM approved, no advanced-clinical SM, no structure-with-ligand, no druggable family)
• Antibody (AB): Advanced Clinical = true (RG7356); UniProt cell-surface localization = high confidence; no approved antibody
• PROTAC (PR): Database ubiquitination and small-molecule binder = true; no clinical PR
• Other clinical (OC): Phase 1 Clinical = true
• Associated diseases in Open Targets: 1,554 (broadly cancer-dominated)

Aging-context druggability-tier = 2. Rationale: A clinical-stage antibody (RG7356) reached Phase 1 in oncology, establishing basic clinical-stage antibody feasibility. However, no agent has been tested for an aging indication; both major oncology programs (bivatuzumab, RG7356) failed. The therapeutic strategy for aging is HA-MW augmentation acting through CD44 rather than CD44 blockade, which is a different modality altogether. For the aging-context tier, tier 2 (“high-quality clinical-stage probe”) is appropriate — there is pharmacological engagement at the clinical level, but not for aging endpoints and not as an approved drug.

Recency notes (2024–2026)

Per R25, recency search is not strictly required for type: protein pages. Notable recent findings via PubMed (2022–2026, n=210 results):

• He et al. 2026 (PMID 41806573; Biochem Biophys Res Commun) — HYAL2-generated LMW-HA promotes nucleus pulposus cell senescence via CD44/AKT axis in intervertebral disc degeneration. Provides mechanistic support for the LMW-HA → CD44 → senescence feedforward loop at a specific aged tissue context. ⁸
• Takasugi 2023 (most important recent finding; see above) — CD44 longevity-correlation signal across species and ATF6/ER-stress mechanism.
• Ongoing drug-delivery literature uses HA-coated nanoparticles for CD44-targeted tumor delivery; not directly aging-relevant but confirms broad bioavailability of HA as a CD44 ligand.

Limitations and knowledge gaps

• Isoform resolution. Most aging-relevant studies use pan-CD44 antibodies without distinguishing CD44s from CD44v isoforms. Whether the aging-relevant signal (contact inhibition, longevity correlation) is carried by CD44s, CD44v, or both is unresolved. needs-isoform-resolution
• HMW-CD44 cluster-size threshold. The precise HA molecular weight at which CD44 clustering becomes sufficient for NF2/merlin activation is not precisely defined. The NMR operates with HA in the MDa range; human tissue HA is in the 0.2–2 MDa range. Whether physiological human HA can engage this axis at all under homeostatic conditions is unclear.
• Direct oncology CD44 programs consistently failed. Despite multiple Phase 1 attempts (bivatuzumab, RG7356), CD44 has proven undruggable as a direct antibody target at efficacious doses in oncology. Whether HA-MW augmentation (a totally different modality) can avoid these failures in aging contexts is untested. needs-human-replication
• Senescence ambiguity. CD44 upregulation on senescent cells and its role in SASP amplification vs. contact-inhibition-driven senescence create mechanistic tension. The net senescence-promoting vs. senescence-suppressing effect of HMW-HA → CD44 in normal aged tissues is unresolved. no-mechanism
• No human aging intervention data. The full HMW-HA → CD44 → NF2/Hippo → cancer-resistance circuit has been established in NMR cells and partially modeled in nmrHas2 mice; there are no human aging endpoint data for any CD44-targeted or HA-augmentation intervention. needs-human-replication

See also

• hyaluronic-acid — HMW-HA vs LMW-HA biology; HA molecular weight determination; NMR-specific HA accumulation; Zhang 2023 mechanism
• has2 — hyaluronic acid synthase 2; the NMR transgene that elevates HMW-HA
• heterocephalus-glaber — naked mole rat model organism page; NMR HMW-HA biology; review Tian 2013 species-specificity claim in light of Zhang 2023
• hippo-yap-taz — the downstream Hippo pathway activated by CD44-NF2 signaling
• nf-kb — activated by LMW-HA → TLR4 → CD44 crosstalk; master SASP driver
• cellular-senescence — CD44 upregulation on senescent cells; SASP bidirectional relationship
• chronic-inflammation — LMW-HA/CD44/TLR4 feedforward inflammaging loop
• stem-cell-exhaustion — CD44 HCELL glycoform in HSC homing; CD44^hi T-cell skewing with aging
• disabled-adaptive-immunity — CD44^hi:CD44^lo T-cell ratio as an immunosenescence readout
• zhang-2023-nmrhas2-mouse-healthspan — primary source for nmrHas2 mouse healthspan/lifespan data
• cancer-aging-tradeoffs — CD44 as a cancer-resistance mechanism with potential aging tradeoffs

---

Footnotes

1. doi:10.1038/nature12234 · Tian X et al. (Gorbunova V, Seluanov A) · Nature 499(7458):346–349 · 2013 · local PDF verified in a local paper archive · in-vivo + in-vitro; naked mole rat HMW-HA (6–12 MDa) → CD44 → NF2 → early contact inhibition → p16^INK4a; CD44 antibody blockade (not siRNA knockdown) abrogated the early-contact-inhibition phenotype. NMR cells displayed two-fold higher HA affinity than mouse/human cells. Established the core HMW-HA/CD44/NF2 cancer-resistance axis. ↩ ↩² ↩³ ↩⁴ ↩⁵ ↩⁶

2. doi:10.1016/j.celrep.2023.113130 · Takasugi M et al. (Seluanov A, Gorbunova V) · Cell Reports 42(9):113130 · 2023 · gold OA · PDF verified in a local paper archive · in-vitro + cross-species transcriptomics (NMR OPCs vs mouse/rat/human OPCs; fibroblasts of 10 mammalian species; liver of 27 mammalian species; skin of 20 rodents); CD44 expression correlates positively with species maximum lifespan independently of HA abundance in tissue (PGLS, p < 0.0064 in fibroblasts); CD44 localizes to the ER and activates basal ATF6 in an HA-independent manner → ER-stress resistance. CD44 siRNA knockdown reduced tunicamycin survival; CD44 overexpression enhanced it. Identifies longevity-correlation and HA-independent ATF6 mechanism downstream of CD44. needs-replication (in-vitro mechanism; no whole-animal lifespan test of CD44 gain-of-function). ↩ ↩² ↩³

3. doi:10.1038/nrm1004 · Ponta H, Sherman L, Herrlich PA · Nature Reviews Molecular Cell Biology 4(1):33–45 · 2003 · 2323 citations · PDF available in a local paper archive · review; comprehensive CD44 domain structure, isoform biology, HA-binding mechanism, ERM interaction, PTMs, and signaling. Foundational reference for CD44 molecular biology. ↩ ↩² ↩³ ↩⁴

4. doi:10.1038/s41586-023-06463-0 · Zhang Z, Tian X et al. (Seluanov A, Gorbunova V) · Nature 621(7977):196–205 · 2023 · 116 citations · PMC10666664 (OA); PDF not in archive · in-vivo (nmrHas2 C57BL/6 mice, n=84 nmrHas2 + 91 controls); +4.4% median lifespan, +12.2% max lifespan, reduced spontaneous cancer mortality (57% vs 70%), inflammaging suppression, gut barrier preservation. See zhang-2023-nmrhas2-mouse-healthspan. ↩ ↩² ↩³

5. doi:10.1084/jem.20001858 · Termeer CC et al. · Journal of Experimental Medicine 195(1):99–111 · 2002 · 1348 citations · in-vitro (dendritic cells); LMW hyaluronan oligosaccharides activate dendritic cells via TLR4 → p38/MAPK/NF-κB. Established LMW-HA as a DAMP activating innate immunity through TLR4. ↩ ↩²

6. doi:10.1073/pnas.0703478104 · Dalerba P et al. (Clarke MF) · PNAS 104(24):10158–10163 · 2007 · PDF verified in a local paper archive · in-vivo (NOD/SCID xenograft); EpCAM^hi/CD44^+ subset initiates CRC tumors across 6 distinct xenograft lines; EpCAM^low/CD44^− fractions consistently non-tumorigenic (0/5 at 10,000 cells). The tumor-initiating phenotype requires EpCAM^hi co-expression; CD44+ alone not tested as sufficient. Established EpCAM^hi/CD44^+ as the colorectal cancer stem cell marker. ↩

7. doi:10.18632/oncotarget.11098 · Menke-van der Houven van Oordt CW et al. · Oncotarget 7(49):80046–80058 · 2016 · PDF verified in a local paper archive · Phase 1 first-in-human clinical trial (n=65; advanced CD44-expressing solid tumors; June 2011–November 2013); RG7356 recombinant humanized IgG1 anti-CD44 antibody; 21% stable disease (13/61 evaluable; median 12 weeks); no partial or complete responses; DLTs: grade 3 headache (1,500 mg q2w; 1,350 mg qw) and grade 4 febrile neutropenia (2,250 mg q2w); MTD 1,500 mg q2w (qw MTD not defined — early study termination). Established insufficient clinical activity for solid tumor monotherapy. ↩

8. doi:10.1016/j.bbrc.2026.153576 · He S et al. · Biochemical and Biophysical Research Communications · 2026 · PMID 41806573 · in-vitro (nucleus pulposus cells); HYAL2-generated LMW-HA suppresses AKT via CD44, promoting cell senescence and ECM degradation in intervertebral disc degeneration. ↩