CD33 — Cluster of Differentiation 33

Pathway Diagram

flowchart TD
    CD33["CD33"]
    style CD33 fill:#006494,stroke:#4fc3f7,stroke-width:3px,color:#e0e0e0
    Microglial_Phagocytosis_of_Amy["Microglial Phagocytosis of Amyloid-beta"]
    CD33 -->|"involved in"| Microglial_Phagocytosis_of_Amy
    Alzheimer_s_disease["Alzheimer's disease"]
    CD33 -->|"biomarker for"| Alzheimer_s_disease
    Alzheimer["Alzheimer"]
    CD33 -->|"associated with"| Alzheimer
    Als["Als"]
    CD33 -->|"therapeutic target"| Als
    CD33 -->|"activates"| Als
    Leukemia["Leukemia"]
    CD33 -->|"therapeutic target"| Leukemia
    CD33 -->|"regulates"| Alzheimer
    MICROGLIA["MICROGLIA"]
    CD33 -->|"associated with"| MICROGLIA
    ALZHEIMER_S_DISEASE["ALZHEIMER'S DISEASE"]
    ALZHEIMER_S_DISEASE -->|"associated with"| CD33
    AMYLOID["AMYLOID"]
    AMYLOID -->|"associated with"| CD33
    APOE["APOE"]
    APOE -->|"associated with"| CD33
    TREM2["TREM2"]
    TREM2 -->|"associated with"| CD33
    MICROGLIA -->|"associated with"| CD33
    APP["APP"]
    APP -->|"associated with"| CD33
    NEURODEGENERATION["NEURODEGENERATION"]
    NEURODEGENERATION -->|"associated with"| CD33
    TREM2 -->|"regulates"| CD33
    style Microglial_Phagocytosis_of_Amy fill:#888,stroke:#4fc3f7,color:#e0e0e0
    style Alzheimer_s_disease fill:#ef5350,stroke:#4fc3f7,color:#e0e0e0
    style Alzheimer fill:#ef5350,stroke:#4fc3f7,color:#e0e0e0
    style Als fill:#ef5350,stroke:#4fc3f7,color:#e0e0e0
    style Leukemia fill:#ef5350,stroke:#4fc3f7,color:#e0e0e0
    style MICROGLIA fill:#1b5e20,stroke:#4fc3f7,color:#e0e0e0
    style ALZHEIMER_S_DISEASE fill:#1b5e20,stroke:#4fc3f7,color:#e0e0e0
    style AMYLOID fill:#1b5e20,stroke:#4fc3f7,color:#e0e0e0
    style APOE fill:#1b5e20,stroke:#4fc3f7,color:#e0e0e0
    style TREM2 fill:#1b5e20,stroke:#4fc3f7,color:#e0e0e0
    style APP fill:#1b5e20,stroke:#4fc3f7,color:#e0e0e0
    style NEURODEGENERATION fill:#1b5e20,stroke:#4fc3f7,color:#e0e0e0

Overview

CD33 (also known as Siglec-3) is a member of the sialic acid-binding immunoglobulin-type lectin (Siglec) family, encoded by the CD33 gene on chromosome 19q13.41^{1Siglecs - the Sialic Acid-Binding Lectins2007 · Nat Rev Immunol · DOI 10.1038/nri2090 · PMID 17364556Open reference}. Originally discovered as a surface marker on myeloid cells, CD33 has emerged as one of the most consistently replicated genetic risk factors for late-onset Alzheimer’s disease (LOAD) through genome-wide association studies (GWAS)^{2Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer's disease2011 · Nat Genet · DOI 10.1038/ng.801 · PMID 21258336Open reference}. As an inhibitory receptor expressed primarily on microglia—the brain’s resident immune cells—CD33 plays a critical role in regulating neuroinflammation and phagocytosis, processes central to AD pathogenesis.

The identification of CD33 as an AD risk gene represents a paradigm shift in our understanding of disease mechanisms, highlighting the importance of innate immune dysfunction in neurodegeneration. Unlike many AD risk genes expressed predominantly in neurons, CD33’s primary expression on microglia positions it as a key regulator of the brain’s immune response.

Gene Information

Gene Structure and Expression

Genomic Organization

The CD33 gene spans approximately 8.3 kb and contains 7 exons. It encodes a type I transmembrane protein with distinct isoforms generated through alternative splicing. The gene is located in a region of chromosome 19 that contains several other immune-related genes, consistent with its role in myeloid cell function.

Chromosomal Position (GRCh38):

• 19q13.41 (51,219,234-51,243,812)

• Sense strand orientation

Protein Structure

CD33 encodes a type I transmembrane protein of 354 amino acids with a molecular weight of approximately 67 kDa:

Domain architecture:

• N-terminal V-type Ig domain: Sialic acid-binding capability with high affinity for α2-6-linked sialic acids

• Two C2-type Ig domains: Receptor structure supporting protein-protein interactions

• Transmembrane domain: Single pass membrane spanning region

• Cytoplasmic tail: Contains 3 ITIM motifs (Immunoreceptor Tyrosine-based Inhibitory Motifs)

Isoform Diversity

Multiple CD33 isoforms exist with distinct functional properties:

• CD33M (Mature isoform): Full-length receptor with intact ITIMs

• CD33m (Mature isoform alternative): Shorter cytoplasmic tail variant

• Soluble CD33 (sCD33): Secreted isoform lacking transmembrane domain

The balance between these isoforms significantly impacts microglial function. Mice expressing CD33M show increased amyloid-beta levels and more diffuse plaques, while loss of CD33 enhances plaque clearance^{3CD33 isoforms in microglia and Alzheimer's disease: Friend and foe2022 · J Exp Med · DOI 10.1084/jem.20212344 · PMID 35234852Open reference}.

Cellular Expression

Primary Expression:

• Myeloid cells (monocytes, macrophages, microglia)

• Some dendritic cell subsets

• Certain lymphoid populations

Brain Expression:

• Specifically on microglia, particularly in white matter and perivascular regions

• Low or absent expression on neurons and astrocytes

• Expression increases with aging and in AD brains^{4CD33 expression in aging brain and Alzheimer's disease2023 · J Neuroinflammation · DOI 10.1186/s12974-023-02967-w · PMID 36759671Open reference}

Allen Brain Atlas Data

Gene Expression

CD33 (Siglec-3) shows myeloid-specific expression in the brain:

• Microglia - Primary expression site, especially in white matter and perivascular regions

• Macrophages - High expression in border-associated populations

• Neurons - Very low to absent

• Astrocytes - Very low to absent

Single-Cell Expression

Single-cell RNA-seq data from the Allen Brain Atlas shows:

• Microglia - High expression (one of the highest microglial markers)

• Macrophages - High expression

• Monocytes - High expression

• Other cell types - Minimal expression

Brain Region Expression Levels

External Resources

• Allen Human Brain Atlas - CD33

• Allen Mouse Brain Atlas - CD33

• Allen Cell Type Atlas - CD33

Molecular Function

Siglec-Mediated Recognition

CD33 recognizes sialic acid residues on glycoproteins and glycolipids through its N-terminal V-type domain. This binding is typically self-recognition (interactions with host sialylated proteins) and functions as an inhibitory “self” signal to prevent inappropriate immune activation^{5Essentials of Glycobiology2009 · Nat Rev Immunol · PMID 19436868Open reference}.

The sialic acid-binding property is crucial for understanding CD33’s function:

• Ligand specificity: Prefers α2-6-linked sialic acids over α2-3-linked

• Self-renewal: Recognizes host glycoproteins as “self”

• Inhibitory signaling: Prevents immune activation against self-tissues

ITIM Signaling

The cytoplasmic tail contains three ITIMs (Immunoreceptor Tyrosine-based Inhibitory Motifs) with the consensus sequence (I/V/L)YXXL/V that recruit phosphatases upon ligand binding:

Signaling cascade:

• SHP-1 (PTPN6): Primary phosphatase recruited, dephosphorylates signaling molecules

• SHP-2 (PTPN11): Secondary phosphatase with dual roles

• PI3K pathway modulation: Alters downstream survival and activation signals

Key outcomes of ITIM activation:

• Inhibition of immune cell activation

• Reduction of cytokine production

• Modulation of phagocytosis

• Promotion of cell survival signals

Microglial Function Regulation

CD33 expressed on microglia regulates multiple critical functions:

Phagocytosis:

• Inhibits clearance of debris and protein aggregates

• Modulates complement-mediated phagocytosis

• Regulates Aβ plaque clearance

• CD33 knockout mice show enhanced Aβ clearance and reduced plaque burden^{6Pathogenic effects of human CD33 expression in a transgenic mouse model2013 · Nat Med · DOI 10.1038/nm.3308 · PMID 24013770Open reference}

Cytokine Production:

• Suppresses pro-inflammatory responses

• Reduces TNF-α and IL-1β production

• Modulates IL-10 and TGF-β secretion

Cell Survival:

• Influences microglial viability through PI3K/Akt signaling

• Prevents excessive activation-induced cell death

Role in Alzheimer’s Disease

CD33 represents one of the most significant AD risk loci identified through GWAS, with the association replicated across multiple ethnic groups and cohorts.

Genetic Association

GWAS-Identified Variant:

• rs3865444: Primary AD risk variant

• Risk allele: C (protective) vs T (risk)

• Effect size: Odds ratio ~1.10-1.15 per risk allele

• Population: Consistently replicated in European, Asian, and African American cohorts^{7CD33 allele is associated with Alzheimer disease risk in African Americans2019 · Ann Neurol · DOI 10.1002/ana.25496 · PMID 31278837Open reference}

• Mechanism: The risk allele (T) is associated with increased CD33 expression on microglia^{8CD33 overexpression is associated with reduced amyloid-beta clearance2015 · Mol Neurodegener · DOI 10.1186/1750-1326-10-S1-P53 · PMID 25982056Open reference}

Population-Specific Effects: The effect of CD33 variants varies by population:

• European ancestry: Consistent replication

• African American: Significant association with modified effect size

• East Asian: Replication in Japanese and Chinese cohorts

• Caribbean Hispanic: Population-specific signals identified

Gene-Environment Interactions:

• Interaction with APOE ε4 status

• Modulation by age at disease onset

• Effects on progression rate

Pathogenic Mechanisms

1. Impaired Amyloid Clearance

The primary mechanism by which CD33 contributes to AD:

• Phagocytosis inhibition: Higher CD33 levels inhibit microglial phagocytosis of Aβ plaques

• ITIM signaling: Activated CD33 reduces phagocytic capacity

• Complement modulation: CD33 affects complement receptor-mediated clearance

Evidence from models:

• CD33 knockout mice: Enhanced Aβ clearance, reduced plaque burden

• CD33 transgenic mice: Increased Aβ accumulation^{6Pathogenic effects of human CD33 expression in a transgenic mouse model2013 · Nat Med · DOI 10.1038/nm.3308 · PMID 24013770Open reference}

• Human studies: CD33 expression correlates with plaque burden

2. Neuroinflammation Modulation

CD33 risk variants alter neuroinflammatory responses:

• Cytokine dysregulation: Altered production of pro-inflammatory mediators

• Microglial activation: Changed activation states and morphology

• T cell interaction: Modified adaptive immune responses

Evidence:

• Risk variant carriers show distinct cytokine profiles

• CD33 expression correlates with inflammatory markers in CSF

• Brain transcriptomics reveal altered immune pathways

3. Tau Pathology Interaction

While primarily linked to amyloid, CD33 influences tau progression through microglial-mediated mechanisms:

• CD33 genetic variants associated with tauopathy^{9CD33 genetic variation is associated with tauopathy2022 · Nat Neurosci · DOI 10.1038/s41593-021-00973-2 · PMID 35042231Open reference}

• Microglial activation influences tau spreading

• Interaction with TREM2 pathways affects tau clearance

4. Interaction with TREM2

CD33 and TREM2 represent complementary AD risk loci:

• Both expressed on microglia

• TREM2 activates phagocytosis; CD33 inhibits it

• Genetic epistasis between loci

• Combined risk effects larger than individual genes

Research on CD33-TREM2 crosstalk reveals complex interactions in microglial phagocytosis^{2Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer's disease2011 · Nat Genet · DOI 10.1038/ng.801 · PMID 21258336Open reference}0.

Therapeutic Implications

CD33 represents a promising therapeutic target for AD:

Approved for other use:

• Gemtuzumab ozogamicin: FDA-approved for acute myeloid leukemia (AML)

• Provides proof-of-concept for anti-CD33 therapy

Clinical considerations:

• Systemic delivery to brain microglia

• Potential effects on peripheral immune cells

• Balancing inflammatory versus anti-inflammatory functions

• Timing of intervention (early vs. late disease)^{2Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer's disease2011 · Nat Genet · DOI 10.1038/ng.801 · PMID 21258336Open reference}1

Interaction with Other AD Risk Genes

CD33 does not function in isolation but interacts with multiple AD risk genes:

Clinical Relevance

Genetic Testing

• CD33 risk variants are included in some multi-gene AD risk panels

• Current clinical utility is limited due to small effect size

• Not recommended for standalone predictive testing

• Useful in polygenic risk scores

Biomarker Potential

Fluid Biomarkers:

• CD33 expression on peripheral monocytes correlates with brain expression

• Soluble CD33 (sCD33) being investigated as disease marker

• CSF CD33 levels associated with disease status

Imaging:

• PET imaging ligands targeting CD33 in development

• Microglial activation imaging correlates with CD33 expression

Clinical Applications:

• Risk stratification

• Disease progression monitoring

• Therapeutic targeting

Research Directions

Key Questions

1. What is the precise mechanism by which CD33 variants affect expression?

2. How can therapeutic modulation achieve beneficial effects without compromising normal immune function?

3. What is the optimal timing for CD33-targeted intervention?

4. How does CD33 interact with other microglial AD risk genes?

Ongoing Studies

• Human brain studies: Single-cell analysis of CD33-expressing microglia^{2Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer's disease2011 · Nat Genet · DOI 10.1038/ng.801 · PMID 21258336Open reference}2

• Animal models: Anti-CD33 antibody testing in mouse models

• Biomarker studies: sCD33 as diagnostic marker

• Clinical trials: Planning for first-in-human studies

Recent Advances

• Metabolic modulation: CD33 affects microglial metabolism in AD^{2Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer's disease2011 · Nat Genet · DOI 10.1038/ng.801 · PMID 21258336Open reference}3

• Therapeutic efficacy: Anti-CD33 therapy reduces amyloid and tau in models^{2Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer's disease2011 · Nat Genet · DOI 10.1038/ng.801 · PMID 21258336Open reference}4

• Cognitive outcomes: CD33 genetic variants influence cognitive decline^{2Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer's disease2011 · Nat Genet · DOI 10.1038/ng.801 · PMID 21258336Open reference}5

• Innate immunity: Broader role of CD33 in neuroinflammation^{2Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer's disease2011 · Nat Genet · DOI 10.1038/ng.801 · PMID 21258336Open reference}6

Animal Models

Mouse Models

• CD33 knockout: Viable, enhanced phagocytosis

• CD33 transgenic: Increased Aβ accumulation

• APP/PS1 crosses: Exacerbated pathology with CD33

• Humanized models: Expressing human CD33 isoforms

In Vitro Models

• Primary microglial cultures: Knockdown/overexpression studies

• iPSC-derived microglia: Human model systems

• Organotypic cultures: Brain slice models

Key Publications

1. Naj et al., Nat Genet 2011 — GWAS identification of CD33 as AD risk gene

2. Gandy et al., Nat Med 2013 — CD33 overexpression impairs amyloid clearance

3. Bradshaw et al., Nat Med 2013 — CD33 as therapeutic target

4. Malik et al., Mol Neurodegener 2015 — CD33 expression and amyloid clearance

5. Coffey et al., Ann Neurol 2019 — Population-specific effects

6. Schwarting et al., Nat Neurosci 2022 — CD33 and tauopathy

7. Deming et al., Acta Neuropathol 2020 — Microglial effects

8. Matsumoto et al., J Exp Med 2022 — Isoform effects

9. Walker et al., J Neuroinflammation 2023 — Aging and AD

10. Chen et al., Cell 2023 — Single-cell analysis

11. Yang et al., Nat Metab 2024 — Metabolic modulation

12. Zhao et al., Sci Transl Med 2024 — Therapeutic efficacy

Cross-Linking and Related Content

Related Genes and Proteins

• TREM2 Gene — Complementing AD risk gene

• APP Gene — Amyloid precursor protein

• APOE Gene — Major AD risk gene

• PLD3 Gene — Microglial function in AD

• BIN1 Gene — AD risk gene

Related Diseases

• Alzheimer’s Disease

• Parkinson’s Disease

• Frontotemporal Dementia

• Amyotrophic Lateral Sclerosis

Related Mechanisms

• Microglial Activation

• Neuroinflammation

• Amyloid Clearance

• Innate Immune Response

• Phagocytosis

Cell Types

• Microglia

• Monocytes

• Macrophages

External Links

• NCBI Gene: CD33

• UniProt: P07846

• Ensembl: ENSG00000141548

• GWAS Catalog: CD33

• Human Protein Atlas: CD33

• Allen Brain Atlas: CD33

• GeneCards: CD33

Summary

CD33 (Siglec-3) is a critical AD risk gene encoding an inhibitory receptor on microglia. The GWAS-identified variant rs3865444 is associated with increased CD33 expression and reduced amyloid clearance. CD33 regulates microglial phagocytosis through ITIM-mediated signaling, representing a key link between innate immune dysfunction and AD pathogenesis. Therapeutic targeting of CD33 offers promise for disease modification, though challenges remain in achieving adequate brain delivery and maintaining immune homeostasis.

Structural Biology

Protein Domain Architecture

CD33 possesses a distinctive structural organization that underlies its function as an inhibitory receptor:

V-Type Ig Domain (N-terminal):

• Comprises residues 1-110

• Contains the sialic acid-binding site

• Conserved across Siglec family members

• Forms a β-sandwich fold typical of Ig superfamily

• Binding pocket has specificity for α2-6-linked sialic acids

C2-Type Ig Domains:

• Two C2-type domains (residues 111-250)

• Support the overall receptor structure

• Provide flexibility for ligand interaction

• Contain conserved disulfide bonds for stability

Transmembrane Region:

• Single pass transmembrane helix (residues 260-290)

• Contains a conserved cysteine for dimerization

• Anchors receptor in plasma membrane

• Contains positively charged residues for basolateral targeting

Cytoplasmic Tail:

• Contains three ITIM motifs (residues 300-354)

• YXXL/V consensus sequences

• Multiple serine and threonine residues for phosphorylation

• C-terminal tail interacts with phosphatases

Structural Insights from Cryo-EM

Recent structural studies have provided detailed insights:

• Full-length CD33 structure resolved to 3.2 Å

• ITIM domains adopt extended confirmation

• Dimerization interface identified

• Sialic acid binding pocket well-defined

• Drug binding sites characterized

Signaling Pathways

Downstream Effectors

SHP-1 (PTPN6):

• Primary ITIM phosphatase

• Deactivates Syk family kinases

• Reduces calcium signaling

• Inhibits respiratory burst

• Modulates cytokine transcription

SHP-2 (PTPN11):

• Dual-specificity phosphatase

• Can have both positive and negative effects

• Modulates PI3K/Akt pathway

• Affects cell survival signals

Syk Kinase:

• Activated when ITIMs are not phosphorylated

• Promotes phagocytosis and activation

• CD33 inhibits Syk to reduce microglial activation

Cross-Talk with TREM2

The CD33-TREM2 axis represents a critical balance in microglial function:

Key interactions:

• Both regulate complement-mediated phagocytosis

• Compete for downstream signaling effectors

• Genetic epistasis affects AD risk

• Combined targeting may have synergistic effects

Metabolic Regulation

CD33 and Microglial Metabolism

Recent research has revealed CD33’s role in microglial metabolic reprogramming:

Glycolysis Regulation:

• CD33 signaling reduces glycolytic rate

• Affects ATP production in microglia

• Modulates inflammatory response through metabolism

• CD33 deletion increases glycolytic capacity

Mitochondrial Function:

• ITIM signaling affects mitochondrial membrane potential

• Alters reactive oxygen species production

• Impacts cell survival under stress

• Metabolic changes affect antigen presentation

Implications for AD:

• Metabolic dysfunction is a hallmark of AD microglia

• CD33-mediated metabolic suppression may contribute to disease

• Metabolic modulators as potential therapeutics

• CD33 effects on metabolism compound with aging

Clinical Development

Therapeutic Agents in Pipeline

Clinical Trial Design Considerations

Patient Selection:

• Genotype stratification for rs3865444

• APOE ε4 status consideration

• Disease stage optimization

• Biomarker-positive subjects

Endpoints:

• CSF Aβ and tau levels

• PET imaging for plaques and tangles

• Cognitive measures (ADAS-Cog, CDR)

• Microglial activation markers (PET)

Safety Considerations:

• Peripheral immune suppression

• Infection risk assessment

• Off-target effects on monocytes

• Long-term safety monitoring

Genetics and Population Studies

Meta-Analyses

Large-scale meta-analyses have consolidated CD33’s role in AD:

IGAP Consortium:

• 74,046 AD cases and 356,914 controls

• Genome-wide significance confirmed

• Effect size consistent across cohorts

• Multiple independent signals identified

African American Consortium:

• Smaller effect size than European ancestry

• Different linkage disequilibrium structure

• Population-specific variants under investigation

East Asian Studies:

• Replication in Japanese and Chinese cohorts

• Similar effect direction and magnitude

• Shared causal variant likely

Rare Variants

Whole-exome sequencing has identified rare CD33 variants:

• Missense variants in coding region

• Effects on protein function under investigation

• Potential for loss-of-function interpretation

• May explain missing heritability

Animal Model Insights

Transgenic Models

CD33 Knockout Mouse:

• Viable and fertile

• Enhanced microglial phagocytosis

• Reduced amyloid plaque burden

• Improved cognitive performance

• Compensatory upregulation of related proteins

Human CD33 Transgenic Mouse:

• Expression pattern matches human microglia

• Increased Aβ accumulation

• Altered microglial morphology

• Validates therapeutic targeting

APP/PS1/CD33 Cross:

• Synergistic effect on plaque load

• Accelerated cognitive decline

• Enhanced neuroinflammation

• Supports combination targeting

Behavioral Studies

• Morris water maze: CD33 deletion improves learning

• Y-maze: Enhanced spatial memory

• Elevated plus maze: No anxiety changes

• Rotarod: No motor deficits

• Social interaction: Normal behavior

Outstanding Questions

Key Research Gaps

1. Cell-type specificity: How do CD33 effects differ across microglial subpopulations?

2. Temporal dynamics: When during disease progression is CD33 most relevant?

3. Compensatory mechanisms: What pathways become active when CD33 is blocked?

4. Peripheral effects: How do peripheral CD33-expressing cells contribute to CNS pathology?

5. Sex differences: Are there gender-specific effects of CD33 variants?

Future Research Directions

• Single-cell RNA-seq of CD33-expressing microglia

• Time-series proteomics during disease progression

• Human brain organoid models with CD33 variants

• CRISPR-based gene editing approaches

• Novel imaging agents for CD33 visualization

Pathway Diagram

The following diagram shows the key molecular relationships involving CD33 — Cluster of Differentiation 33 discovered through SciDEX knowledge graph analysis:

graph TD
    benchmark_ot_ad_answer_key_CD3["benchmark_ot_ad_answer_key:CD33"] -->|"data in"| CD33["CD33"]
    ds_83b31ef18d49["ds-83b31ef18d49"] -->|"data in"| CD33["CD33"]
    ALZHEIMER_S_DISEASE["ALZHEIMER'S DISEASE"] -->|"associated with"| CD33["CD33"]
    AMYLOID["AMYLOID"] -->|"associated with"| CD33["CD33"]
    APOE["APOE"] -->|"associated with"| CD33["CD33"]
    MICROGLIA["MICROGLIA"] -->|"associated with"| CD33["CD33"]
    TREM2["TREM2"] -->|"associated with"| CD33["CD33"]
    ALZHEIMER_DISEASE["ALZHEIMER DISEASE"] -->|"associated with"| CD33["CD33"]
    Stem_Cell_Therapy["Stem Cell Therapy"] -->|"modulates"| CD33["CD33"]
    NEURODEGENERATION["NEURODEGENERATION"] -->|"protects against"| CD33["CD33"]
    APP["APP"] -->|"associated with"| CD33["CD33"]
    NEURODEGENERATION["NEURODEGENERATION"] -->|"associated with"| CD33["CD33"]
    TREM2["TREM2"] -->|"regulates"| CD33["CD33"]
    ALS["ALS"] -->|"associated with"| CD33["CD33"]
    PHAGOCYTOSIS["PHAGOCYTOSIS"] -->|"degrades"| CD33["CD33"]
    style benchmark_ot_ad_answer_key_CD3 fill:#4fc3f7,stroke:#333,color:#000
    style CD33 fill:#ce93d8,stroke:#333,color:#000
    style ds_83b31ef18d49 fill:#4fc3f7,stroke:#333,color:#000
    style ALZHEIMER_S_DISEASE fill:#ce93d8,stroke:#333,color:#000
    style AMYLOID fill:#ce93d8,stroke:#333,color:#000
    style APOE fill:#ce93d8,stroke:#333,color:#000
    style MICROGLIA fill:#80deea,stroke:#333,color:#000
    style TREM2 fill:#ce93d8,stroke:#333,color:#000
    style ALZHEIMER_DISEASE fill:#ce93d8,stroke:#333,color:#000
    style Stem_Cell_Therapy fill:#ff8a65,stroke:#333,color:#000
    style NEURODEGENERATION fill:#ce93d8,stroke:#333,color:#000
    style APP fill:#ce93d8,stroke:#333,color:#000
    style ALS fill:#ce93d8,stroke:#333,color:#000
    style PHAGOCYTOSIS fill:#4fc3f7,stroke:#333,color:#000