Toll-Like Receptor

Toll-like receptors (TLRs) are an evolutionary conserved family of germline-encoded, type I transmembrane pattern recognition receptors (PRRs) that serve as primary molecular sentinels of the innate immune system. Expressed prominently on sentinel leukocytes—including macrophages, conventional and plasmacytoid dendritic cells, neutrophils, mast cells, and B cells—as well as on mucosal epithelial and vascular endothelial cells, TLRs detect conserved microbial structures termed pathogen-associated molecular patterns (PAMPs) alongside endogenous host molecules released from necrotic or stressed tissues, termed damage-associated molecular patterns (DAMPs). Humans express 10 functional TLRs (TLR1 through TLR10), whereas mice express 12 (TLR1–TLR9 and TLR11–TLR13, with murine Tlr10 being a non-functional pseudogene).

The conceptual foundation for TLR biology was established in 1989 when Charles Janeway Jr. proposed that innate immunity does not merely respond to non-specific inflammation, but relies on germline-encoded receptors evolved to recognize invariant microbial macromolecules essential for pathogen survival (such as bacterial cell-wall components or viral nucleic acids) and thereby license adaptive T- and B-cell activation. In 1996, Jules Hoffmann and colleagues demonstrated that the Drosophila melanogaster Toll receptor—originally identified by Christiane Nüsslein-Volhard for governing dorsoventral embryonic polarity—was required to induce the antifungal peptide drosomycin against Aspergillus fumigatus. Shortly thereafter, Ruslan Medzhitov and Charles Janeway cloned a human homolog (now designated TLR4) capable of activating NF-κB and inducing the costimulatory molecule B7 (CD80), and in 1998 Bruce Beutler used positional cloning in endotoxin-resistant C3H/HeJ and C57BL/10ScCr mice to prove that TLR4 is the long-sought mammalian signaling receptor for Gram-negative bacterial lipopolysaccharide (LPS), discoveries recognized with the 2011 Nobel Prize in Physiology or Medicine.

Ectodomain Leucine-Rich Repeats and the Cytoplasmic TIR Domain

Every Toll-like receptor is a type I integral membrane glycoprotein organized into three modular domains:

1. Ligand-Binding Ectodomain (ECD): Measuring ~550 to 800 amino acids, the N-terminal extracellular (or endolumenal) domain consists of 19 to 25 tandem leucine-rich repeat (LRR) motifs. Each LRR is 24 to 29 residues long and contains the conserved consensus motif LxxLxLxxNxL, in which hydrophobic leucine side chains pack into the interior core while parallel β-strands align along the concave inner surface and variable α-helices or loops form the convex outer surface. Together, the tandem LRRs fold into a characteristic solenoid horseshoe (arc) structure capped at the N- and C-termini by disulfide-stabilized LRRNT and LRRCT cysteine clusters. Ligand binding to the lateral convex or ascending faces of two TLR ectodomains stabilizes an M-shaped dimeric conformation (either a homodimer or a heterodimer), bringing their C-terminal juxtamembrane regions within ~20 Å of one another.
2. Transmembrane Helix: A single hydrophobic α-helix of ~20 amino acids spans the lipid bilayer and transmits the ligand-induced dimerization of the ectodomains across the membrane.
3. Cytoplasmic TIR Domain: Homologous to the intracellular region of the interleukin-1 receptor (IL-1R), the ~150-residue Toll/IL-1 receptor (TIR) domain adopts a globular α/β fold consisting of a central five-stranded parallel β-sheet surrounded by five α-helices. A conserved surface loop between the second β-strand and second α-helix—the BB loop (where a single Pro712His point mutation in C3H/HeJ mice abolishes TLR4 signaling)—mediaphores homotypic protein–protein interactions that recruit TIR-domain-containing cytosolic sorting and signaling adaptors.

Cell-Surface versus Endosomal Localization

The 10 human TLRs are spatially segregated between the plasma membrane and intracellular endolysosomal compartments, an evolutionary adaptation that matches each receptor to the anatomical compartment where its target ligand is encountered while preventing accidental activation by host self-nucleic acids:

• Cell-Surface TLRs (TLR1, TLR2, TLR4, TLR5, TLR6, and TLR10): Positioned on the plasma membrane, these receptors detect exposed structural lipids, lipoproteins, and proteins of extracellular bacteria, fungi, and parasites. TLR2 forms heterodimers with TLR1 to recognize triacylated lipopeptides (from Gram-negative bacteria and mycobacteria) or with TLR6 to recognize diacylated lipopeptides (such as mycoplasmal MALP-2) and Gram-positive bacterial lipoteichoic acid and peptidoglycan, as well as fungal β-glucan (zymosan) in cooperation with Dectin-1. TLR4 complexes with the soluble accessory protein MD-2 (LY96) to bind the hexa-acylated lipid A moiety of Gram-negative LPS, delivered by LPS-binding protein (LBP) and the glycosylphosphatidylinositol (GPI)-anchored co-receptor CD14. TLR5 recognizes a conserved D1 domain motif on monomeric bacterial flagellin, whereas human TLR10 acts primarily as a modulatory receptor that heterodimerizes with TLR2 to dampen inflammatory cytokine production.
• Endosomal TLRs (TLR3, TLR7, TLR8, and TLR9): Sequestered within early endosomes, late endosomes, and lysosomes, these receptors detect viral and microbial nucleic acids liberated only after virions or bacteria are endocytosed and digested by acid hydrolases. Synthesized in the endoplasmic reticulum (ER), endosomal TLRs require the multi-pass transmembrane chaperone UNC93B1 (alongside gp96 and PRAT4A) to exit the ER via COPII vesicles and traffic to endolysosomes. Once inside the acidic endolysosome (pH 4.5–6.0), acid-dependent cathepsins and asparagine endopeptidase (AEP) proteolytically cleave the Z-loop of the ectodomains of TLR3, TLR7, TLR8, and TLR9—a mandatory processing step that enables functional dimerization while ensuring that any receptor mislocalized to the neutral cell surface cannot respond to extracellular host DNA or RNA released from dying cells.

The MyD88–Myddosome Pathway

Ligand-induced dimerization of the cytoplasmic TIR domains recruits specific combinations of four TIR-domain-containing adaptor proteins: MyD88 (Myeloid Differentiation Primary Response 88), TIRAP (also called Mal), TRIF (TICAM-1), and TRAM (TICAM-2). All TLRs except TLR3 utilize MyD88. At the plasma membrane, TLR2 and TLR4 require the bridging adaptor TIRAP, which contains an N-terminal phosphatidylinositol 4,5-bisphosphate (PIP₂)-binding domain that recruits MyD88 to lipid rafts.

MyD88 possesses a C-terminal TIR domain and an N-terminal death domain (DD). Upon receptor engagement, MyD88 oligomers nucleate a left-handed helical supramolecular signaling complex called the Myddosome, assembling 6 to 8 MyD88 molecules with 4 molecules of the serine/threonine kinase IRAK4 and 4 molecules of IRAK1 (or IRAK2). Proximity within the Myddosome enables IRAK4 to autophosphorylate and subsequently phosphorylate IRAK1, which dissociates from the receptor complex to interact with the RING-domain E3 ubiquitin ligase TRAF6. Together with the E2 conjugating enzyme complex Ubc13–Uev1A, TRAF6 catalyzes the synthesis of lysine-63 (K63)-linked polyubiquitin chains (and linear M1-linked ubiquitin chains via the LUBAC complex). These non-degradative ubiquitin scaffolds recruit the TAK1 kinase complex (TAK1, TAB1, TAB2/3) and the IKK complex (IKKα, IKKβ, NEMO). TAK1 phosphorylates IKKβ—which degrades IκBα to liberate NF-κB (p50/p65)—and simultaneously phosphorylates MAP kinase kinases (MKK3/6 and MKK4/7) to activate p38 and JNK, driving AP-1 (c-Fos/c-Jun) transcription of pro-inflammatory cytokines (TNF, IL1B, IL6, IL12B).

In plasmacytoid dendritic cells (pDCs), endosomal TLR7 and TLR9 utilize MyD88 in a specialized bifurcation: the Myddosome directly recruits TRAF6, IRAK1, IKKα, and the transcription factor IRF7. Phosphorylation of IRF7 by IRAK1 and IKKα triggers its nuclear translocation, enabling pDCs to secrete up to 1,000-fold more Type I interferon (IFN-α) than any other blood leukocyte during systemic viral infections.

The TRIF-Dependent Pathway and TLR4 Endocytosis

In contrast, TLR3 signals exclusively through TRIF, while TLR4 uniquely activates both pathways in sequential spatiotemporal waves. Following initial engagement of LPS at the plasma membrane and activation of TIRAP–MyD88 signaling, CD14 and the spleen tyrosine kinase (Syk)–PLCγ2 pathway trigger dynamin-dependent endocytosis of the TLR4–MD-2–LPS complex into early endosomes. As PIP₂ is hydrolyzed, TIRAP dissociates and the myristoylated bridging adaptor TRAM binds the endosomal membrane to recruit TRIF.

TRIF serves as a platform for two downstream branches: it binds TRAF6 and RIPK1 (via a C-terminal RHIM motif) to sustain late-phase NF-κB activation, and it recruits TRAF3 along with the non-canonical IKK-related kinases TBK1 (TANK-binding kinase 1) and IKKε. TBK1 phosphorylates a C-terminal serine cluster on Interferon Regulatory Factor 3 (IRF3), causing IRF3 to homodimerize, translocate to the nucleus, and partner with NF-κB and ATF2/c-Jun on the enhanceosome of the IFNB1 promoter to transcribe IFN-β. Together, TLR-induced cytokines and Type I interferons upregulate MHC class II and the costimulatory ligands CD80 and CD86 on dendritic cells, licensing naive T-cell priming. This functional bridge is exploited clinically in modern vaccine adjuvants—such as monophosphoryl lipid A (MPLA), a detoxified TLR4 agonist used in Shingrix and Cervarix, and synthetic CpG oligodeoxynucleotides (CpG 1018), a TLR9 agonist used in the Heplisav-B hepatitis B vaccine—as well as topical imiquimod, a TLR7 agonist used to treat basal cell carcinoma and genital warts.

How does the innate immune system prevent endosomal TLR7 and TLR9 from triggering lupus-like autoimmunity against host RNA and DNA?

Three layers of checkpoint control prevent self-nucleic acid recognition: first, self-DNA and RNA released by apoptotic cells are rapidly degraded in extracellular fluid by serum DNase I and RNases before reaching endosomes; second, UNC93B1 restricts TLR7 and TLR9 activation strictly to acidified endolysosomes where cathepsin/AEP cleavage occurs; and third, UNC93B1 recruits syntenin-1 to limit TLR7 signaling, whereas gain-of-function TLR7 mutations that bypass this restraint cause human systemic lupus erythematosus (SLE).

Why is monophosphoryl lipid A (MPLA) safe as a human vaccine adjuvant whereas native bacterial lipid A causes lethal septic shock?

Native Escherichia coli or Salmonella lipid A possesses two glucosamine sugars phosphorylated at both the 1 and 4′ positions and six acyl chains, which strongly triggers both plasma-membrane TIRAP–MyD88 pro-inflammatory signaling and endosomal TRAM–TRIF signaling. Chemical removal of the 1-phosphate group yields monophosphoryl lipid A (MPLA), which alters TLR4–MD-2 dimer geometry to bias signaling toward the TRIF–IRF3 adaptive-licensing pathway while reducing MyD88-driven TNF-α and IL-1β toxicity by more than 100-fold.

Why do humans with inherited MyD88 or IRAK4 deficiency suffer primarily from pyogenic bacterial infections in infancy rather than broad viral susceptibility?

Because human antiviral immunity relies heavily on TLR3–TRIF signaling and cytosolic nucleic acid sensors (RIG-I, MDA5, and cGAS–STING) that operate independently of MyD88, patients with autosomal recessive MyD88 or IRAK4 deficiency handle most common viral infections normally. However, their macrophages and neutrophils cannot mount rapid TLR2/TLR4/IL-1R pro-inflammatory cytokine responses to encapsulated pyogenic bacteria such as Streptococcus pneumoniae and Staphylococcus aureus until adaptive antibody responses mature in late childhood.

1. Medzhitov, R., Preston-Hurlburt, P., & Janeway, C. A. Jr. (1997). "A human homologue of the Drosophila Toll protein signals activation of adaptive immunity." Nature, 388(6640), 394–397. https://doi.org/10.1038/41131
2. Poltorak, A., He, X., Smirnova, I., Liu, M. Y., Van Huffel, C., Du, X., Birdwell, D., Alejos, E., Silva, M., Galanos, C., Freudenberg, M., Ricciardi-Castagnoli, P., Layton, B., & Beutler, B. (1998). "Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: Mutations in Tlr4 gene." Science, 282(5396), 2085–2088. https://doi.org/10.1126/science.282.5396.2085
3. Akira, S., Uematsu, S., & Takeuchi, O. (2006). "Pathogen recognition and innate immunity." Cell, 124(4), 783–801. https://doi.org/10.1016/j.cell.2006.02.015
4. Fitzgerald, K. A., & Kagan, J. C. (2020). "Toll-like receptors and the control of immunity." Cell, 180(6), 1044–1066. https://doi.org/10.1016/j.cell.2020.02.041