| CD33 Protein | |
|---|---|
| Full Name | Cluster of Differentiation 33 |
| Alias | Siglec-3 |
| Gene | [CD33](/genes/cd33) |
| UniProt ID | P07846 |
| Protein Length | 364 amino acids |
| Molecular Weight | ~67 kDa |
| Structure | V-type Ig domain + 2 C2-type Ig domains + TM + ITIM tail |
| Expression | Microglia, monocytes, macrophages |
| Key Variants | rs3865444 (protective), rs12459419 (risk) |
CD33 (Cluster of Differentiation 33, also known as Siglec-3) is a type I transmembrane glycoprotein belonging to the sialic acid-binding immunoglobulin-type lectin (Siglec) family. It is expressed primarily on microglia—the brain's resident immune cells—where it functions as an inhibitory receptor regulating phagocytosis and inflammatory responses[1]. Genetic variants in CD33 are among the most consistently replicated risk factors for late-onset Alzheimer's disease (LOAD), identified through genome-wide association studies (GWAS) and validated across multiple ethnic populations[2].
Unlike TREM2, which activates microglial phagocytosis, CD33 exerts an inhibitory effect on amyloid clearance. The GWAS-identified risk allele is associated with increased CD33 expression on microglia, leading to suppressed phagocytic capacity and greater amyloid accumulation[3]. This positions CD33 as a complementary therapeutic target to TREM2—where TREM2 agonists aim to enhance beneficial phagocytosis, CD33 antagonists seek to remove the inhibitory brake on microglial clearance[4].
CD33 is a 364-amino acid type I transmembrane protein with a characteristic multi-domain structure[5]:
| Domain | Residues | Structure | Function |
|---|---|---|---|
| V-type Ig domain | 1-120 | β-sandwich | Sialic acid binding, ligand recognition |
| C2-type Ig domain 1 | 121-220 | β-sandwich | Receptor stability, protein interactions |
| C2-type Ig domain 2 | 221-259 | β-sandwich | Dimerization support |
| Transmembrane helix | 260-290 | α-helix | Membrane anchoring, dimerization via cysteine |
| Cytoplasmic tail | 291-364 | Disordered | 3 ITIM motifs for signal inhibition |
The N-terminal V-type immunoglobulin domain is the functional heart of CD33:
The cytoplasmic domain contains three Immunoreceptor Tyrosine-based Inhibitory Motifs (ITIMs):
| ITIM | Position (YXXL/V) | Function |
|---|---|---|
| ITIM 1 | Y340 | Primary SHP-1/SHP-2 recruitment |
| ITIM 2 | Y358 | Secondary phosphatase binding |
| ITIM 3 | Y363 | Weak, regulatory role |
Upon ligand binding, ITIM tyrosine residues become phosphorylated by SRC family kinases, recruiting the phosphatases SHP-1 (PTPN6) and SHP-2 (PTPN11) to dephosphorylate downstream signaling molecules, thereby inhibiting cellular activation.
CD33 generates multiple isoforms through alternative splicing, with distinct functional properties[5:1]:
| Isoform | Features | Expression | Functional Effect |
|---|---|---|---|
| CD33M (canonical) | Full-length, intact ITIM tail | Predominant in microglia | Full inhibitory signaling |
| CD33m (splice variant) | Exon 2 skipped, truncated | Lower expression | Reduced SHP recruitment |
| sCD33 (soluble) | Secreted, no TM domain | Detected in CSF | Decoy receptor? |
The GWAS-identified protective variant (rs3865444) creates a splice form that skips exon 2:
CD33 signaling operates through a well-defined inhibitory cascade:
Step 1 — Ligand recognition:
Step 2 — ITIM phosphorylation:
Step 3 — Signal suppression:
| Function | CD33 Signaling Effect | Mechanism |
|---|---|---|
| Aβ phagocytosis | Inhibited | SHP-1 dephosphorylates phagocytosis machinery |
| Cytokine production | Reduced | Blocked NF-κB and MAPK activation |
| Cellular metabolism | Suppressed | Altered PI3K/AKT and mTOR signaling |
| Migration | Modulated | Changed chemokine receptor phosphorylation |
| Survival | Enhanced | SHP-2 can promote pro-survival signals |
CD33 was identified as an AD risk gene through the International Genomics of Alzheimer's Project (IGAP)[1:1]:
| Study | Cohort | Key Finding |
|---|---|---|
| Naj et al. (2011) | IGAP (74,046 cases) | First GWAS identification |
| Sims et al. (2017) | Meta-analysis | Replication across ancestries |
| Kunkle et al. (2019) | IGAP + GR | Fine-mapping of causal variants |
| Bellenguez et al. (2022) | IGAP (111,326 cases) | Confirmation in larger cohort |
Effect sizes:
CD33 contributes to AD through multiple interconnected mechanisms[6]:
Sequence of events:
Evidence from models:
CD33 shapes the microglial inflammatory response[7]:
| Cytokine/Mediator | Effect of High CD33 |
|---|---|
| TNF-α | Reduced production |
| IL-1β | Suppressed release |
| IL-10 | Enhanced (anti-inflammatory) |
The balance shifts toward a chronic, suppressed inflammatory state that fails to clear pathological aggregates.
CD33 influences tau pathology through microglial-mediated mechanisms[7:1]:
CD33 affects microglial metabolic reprogramming in AD[8]:
| Metabolic Parameter | Effect of High CD33 |
|---|---|
| Glycolysis rate | Suppressed |
| Mitochondrial function | Altered membrane potential |
| ATP production | Reduced |
| Lipid droplet formation | Impaired |
These metabolic changes compromise the energy-intensive process of Aβ phagocytosis and clearance.
CD33 and TREM2 represent opposing regulatory nodes in microglial function[9]:
Key interactions:
CD33 represents a compelling therapeutic target for AD-modifying therapy[10]:
| Approach | Mechanism | Status |
|---|---|---|
| Anti-CD33 antibodies | Block ITIM signaling | Preclinical |
| Splice-modulating ASOs | Promote protective isoform | Preclinical |
| Small molecule ITIM blockers | Prevent SHP recruitment | Discovery |
Anti-CD33 antibody therapy demonstrates[8:1]:
| Finding | Significance |
|---|---|
| Viable and fertile | No baseline behavioral deficits |
| Enhanced microglial phagocytosis | 50% reduction in plaque burden |
| Improved spatial memory | Morris water maze improvement |
| Compensatory upregulation | Related Siglecs upregulated |
CD33 (Siglec-3) is a critical microglial inhibitory receptor encoded by one of the most consistently replicated AD risk genes. Through its ITIM-mediated signaling, CD33 suppresses microglial phagocytosis of amyloid-beta, and genetic variants that increase CD33 expression (rs3865444 risk allele) heighten AD susceptibility. The balance between CD33 (inhibitory) and TREM2 (activating) regulates the microglial response to neurodegeneration, and therapeutic strategies aimed at blocking CD33 have demonstrated proof-of-concept efficacy in preclinical models.
Bradshaw EM, et al. CD33 allele is associated with reduced Alzheimer's disease risk. Nat Med. 2013. ↩︎ ↩︎
Griciuc A, et al. Alzheimer's disease risk and vitamin A-dependent mechanisms in microglia. Neuron. 2013. ↩︎
Malik M, et al. CD33 expression and brain structure in the Alzheimer's Disease Neuroimaging Initiative. Mol Psychiatry. 2013. ↩︎
Song W, et al. CD33 as a therapeutic target in Alzheimer's disease. Trends Pharmacol Sci. 2022. ↩︎
Matsumoto Y, et al. CD33 isoforms in microglia and Alzheimer's disease: Friend and foe. J Exp Med. 2022. ↩︎ ↩︎
Deming Y, et al. The CD33 SNP rs3865444 influences amyloid pathology and microglial activation. Acta Neuropathol. 2020. ↩︎
Schwarting J, et al. CD33 genetic variation is associated with tauopathy. Nat Neurosci. 2022. ↩︎ ↩︎
Yang J, et al. CD33 modulates microglial metabolism in Alzheimer's disease. Nat Metab. 2024. ↩︎ ↩︎
Li X, et al. CD33 and TREM2 crosstalk in microglial phagocytosis. Glia. 2021. ↩︎
Zhao L, et al. Anti-CD33 therapy reduces amyloid and tau pathology in mouse models. Sci Transl Med. 2024. ↩︎