5.04: B. The Innate Immune System, PAMPs and DAMPs, and Inflammation

Search Fundamentals of Biochemistry

Learning Goals 

(Learning goals written by Claude, Sonnet 4.6, Anthropic)

Pattern Recognition in the Innate Immune System

• Distinguish pathogen-associated molecular patterns (PAMPs) from damage-associated molecular patterns (DAMPs), identify the chemical nature of representative examples of each (LPS, mannose, dsRNA, unmethylated CpG DNA; extracellular ATP, uric acid crystals, heat shock proteins, histones), and explain why recognizing common molecular patterns rather than individual antigens makes the innate immune system broadly effective without requiring adaptation to each new pathogen.
• Describe the structural basis of TLR3 recognition of dsRNA — including the horseshoe-shaped leucine-rich repeat solenoid, the ligand-induced dimerization requiring ≥40–50 bp dsRNA, and the extensive glycosylation — and explain why dsRNA (absent from most host systems) makes an ideal PAMP for viral detection, connecting this to the molecular innovation of modified nucleosides (pseudouridine) in mRNA vaccines to prevent TLR7/8/3-mediated immune activation against the vaccine RNA itself.
• Describe the cGAS-STING and OAS-RNaseL innate immune signaling axes — explaining how cytoplasmic viral dsDNA/dsRNA activates the synthases cGAS and OAS to produce the small-molecule second messengers cGAMP and 2',5'-oligoadenylate, how these bind downstream effectors STING and RNaseL to initiate antiviral responses, and how viruses counteract this pathway by deploying LigT-like phosphodiesterases (identified by FoldSeek structural comparison) that hydrolyze these cyclic nucleotide messengers.

The Inflammasome: Assembly, Mechanism, and Disease Relevance

• Explain the two-signal requirement for NLRP3 inflammasome-mediated cytokine release — Signal 1 (PAMP binding to TLRs → NF-κB activation → transcription of pro-IL-1β, pro-IL-18, and NLRP3 itself) and Signal 2 (DAMP/stress sensing → NLRP3 activation → inflammasome assembly) — and rationalize why two signals are required as a safety mechanism analogous to the two-signal requirement for T-cell activation, preventing inadvertent inflammatory damage.
• Trace the molecular pathway from DAMP sensing to cytokine release: DAMPs (crystals, extracellular ATP, lysosomal rupture) → K⁺ efflux via pannexin-1/P2X7 → NEK7 binding and activation of NLRP3 → pyrin:pyrin recruitment of ASC adaptor → CARD:CARD recruitment of procaspase-1 → autocatalytic activation of caspase-1 → proteolytic release of active IL-1β and IL-18.
• Compare the structural basis and activation mechanisms of NLRC4 (NAIP2:NLRC4) and NLRP3 inflammasomes — contrasting specific PAMP-mediated activation of NAIP2 (binding bacterial rod protein PrgJ, flagellin) that triggers sequential NLRC4 subunit recruitment through complementary electrostatic surfaces into a precise 11-mer ring with the indirect, stress-sensing activation of NLRP3 that responds to a broad range of DAMPs — and explain how CARD domain-mediated proximity activation of procaspase-1 is mechanistically shared between both inflammasome types.
• Explain the domain architecture of the major inflammasome components — including the pyrin, NACHT, LRR, CARD, and BIR domains found in NLRPs, NAIPs, ASC, and caspase-1 — and describe how domain-domain interactions (pyrin:pyrin and CARD:CARD) drive the sequential assembly of the inflammasome supercomplex, connecting domain-mediated protein-protein interactions to the broader principle that multi-domain protein complexes use structured interfaces to achieve regulated assembly.

Recognition and Response in the Innate Immune System

The B and T cell part of the immune system represents the more sophisticated branch called the adaptive immune system. It can be trained to recognize any foreign chemical/cellular species. The other branch of the immune system is the innate immune system. The system recognizes common molecular structures found all many different organisms, so in this branch, there is no need to adapt to each foreign species individually. The adaptive immune response also must be activated by cells of the innate immune system.

The innate immune system recognizes common structural features in viruses and living cells, like bacteria, fungi, and protozoans like amoebas. The cells of the innate system (dendritic cells, macrophages, eosinophils, etc, which we talked about as antigen-presenting calls above) have receptors called Toll-like Receptors 1-10 (TLRs) that recognize the common pathogen-associated molecular patterns (PAMPs), which leads to binding, engulfment, signal transduction, maturation (differentiation), antigen presentation, and cytokine/chemokine release from these cells. Dendritic cells, which reside in the peripheral tissues and act as sentinels, are an example. They can bind PAMPs, which include:

• CHO/Lipids on the bacterial surface (LPS)
• mannose (CHO found in abundance on bacteria,
• yeast dsRNA (from viruses)
• nonmethylated CpG motifs in bacterial DNA

Upon entry into an immune cell, bacterial and viral nucleic acids are recognized by intracellular TLRs. Dendritic cells phagocytize microbial and host cells killed through programmed cell death (apoptosis). During maturation, surface protein expression is altered, allowing cells to leave peripheral tissues and migrate to lymph nodes, where they activate T cells through the antigen-presenting mechanisms described above. They also regulate lymphocyte migration by releasing chemokines. Figure \(\PageIndex{16}\) shows the TLR family, their binding signals, and intracellular adapter proteins that transmit signals into the cell.

Figure \(\PageIndex{16}\): TLR family, their binding signals, and intracellular adapter proteins Ji-Yoon Noh, Suk Ran Yoon, Tae-Don Kim, Inpyo Choi, Haiyoung Jung, "Toll-Like Receptors in Natural Killer Cells and Their Application for Immunotherapy", Journal of Immunology Research, vol. 2020, Article ID 2045860, 9 pages, 2020. https://doi.org/10.1155/2020/2045860. This is an open-access article distributed under the Creative Commons Attribution License

 

Figure \(\PageIndex{16b}\) shows an interactive iCn3D model of the TLR3 human Toll-like receptor 3 (or TLR3).  Only 1 member of the TLR3 dimer is shown.

Figure \(\PageIndex{16b}\): TLR3 human Toll-like receptor 3 (or TLR3).  From Membranome.  https://membranome.org/proteins/4119.  Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/i...tSMzQgqWzbgQg7  

The dsRNA ligand binding causes dimerization of two TLR3 chains. Figure \(\PageIndex{17}\) shows an interactive iCn3D model of the mouse Toll-like receptor three ectodomain (that sticks out into the cytoplasmic space from an internal organelle) complexed with double-stranded RNA (3CIY).  Two TLR ectodomain chains are shown, and the dsRNA is bound between them.

Figure \(\PageIndex{17}\): Mouse Toll-like receptor 3 ectodomain complexed with double-stranded RNA (3CIY) (Copyright; author via source). Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/icn3d/share.html?otC8JXNQe45xBFx47

Double-stranded RNA is found in the life cycles of many viruses, so it makes great sense for evolution to create a binding protein that recognizes this common structure (a PAMP). The TLR3 ectodomains (ECDs) form dimers when the dsRNA is at least 40-50 nucleotides long. The dsRNA is shown in spacefill (cyan and magenta). Note the extensive glycosylation (colored cubes) in the structure. The protein looks like a "horseshoe-shaped solenoid " with lots of beta structure. It has 23 leucine-rich repeats (LRRs) with some conserved asparagines, allowing for extensive hydrogen bonding. One face appears free of carbohydrate residues and may be important in dimerization and function.

---

Recent Updates:  October 27, 2024

Chapter 3.4 discussed how FoldSeek was used to predict the function of many new viral proteins from their 3D structures.  Some appear to inhibit the innate response of the host (either eukaryotes or prokaryotes) to the virus. Host cells have viral sensors that recognize PAMPs (described above).  The recognition by the innate immune system of viral dsDNA and dsRNA in humans can lead to the synthesis of small molecules (messengers) like  2',3'-cyclic GMP-AMP (cGAMP), 2',5'-oligoadenylate (2',5'-OA) by the enzymes cyclic GMP-AMP synthase (cGAS) and oligoadenylate synthase (OAS), respectively.  Similarly, a small molecule messenger, 3',3',cGAMP, is synthesized by the enzyme nucleotidyltransferases (CD-NTases) in prokaryotes.  These enzymes are activated by dsDNA/dsRNA and act as sensors of viral invasion.  Figure \(\PageIndex{18}\) overviews the players in these innate immune system activation pathways.

Figure \(\PageIndex{18}\): Overview of small molecule messengers synthesis responses in the innate immune system activated by viral dsDNA/ds RNA.  Nomburg, J., Doherty, E.E., Price, N. et al. Birth of protein folds and functions in the virome. Nature 633, 710–717 (2024). https://doi.org/10.1038/s41586-024-07809-y.  Creative Commons Attribution 4.0 International License. http://creativecommons.org/licenses/by/4.0/.

In these cases, dsDNA or dsRNA binds to the cGAS, OAS, or CD-NTase synthases.  On binding, the nucleic acid activates the synthesis of these small-molecule messengers.  The small molecules then bind to "downstream" effector proteins like STING (Stimulator of interferon genes protein) and RNaseL (2-5A-dependent ribonuclease) in humans to initiate antiviral immune activity.

The structures of the small "signaling" molecules, 2',5'-oligoadenylate (2',5'-OA) and 2',3'-cyclic GMP-AMP (cGAMP) are shown in Figure \(\PageIndex{19}\) below.

2',3'-cyclic GMP-AMP (cGAMP). Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/i...nnZeMcAYiJL8U6

Figure \(\PageIndex{19}\):  Small signaling molecules synthesized in the innate immune system on viral infection

 

Figure \(\PageIndex{20}\) shows an interactive iCn3D model of mouse cGAS bound to an 18bp DNA and cGAMP (4LEZ). cGAMP is shown in spacefill.

Figure \(\PageIndex{20}\): Mouse cGAS bound to an 18bp DNA and cGAMP (4lez). (Copyright; author via source). Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/i...C3FP4faWa9BgD8

The dimeric cGAS binds two dsDNAs, leading to dimerization and synthase activation.  The spacefill cGAMP is the product of the reaction. 

The product of the synthase, cGAMP, binds to a downstream effector protein, STING, which ultimately leads to the synthesis of proteins involved in the antiviral response. Figure \(\PageIndex{21}\) shows an interactive iCn3D model of a human STING mutant R232K in complex with 2',3'-cGAMP (6Y99)

Figure \(\PageIndex{21}\): human STING mutant R232K in complex with 2',3'-cGAMP (6Y99). (Copyright; author via source). Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/i...zwQ2WBX2UQgzg7

May 8, 2026

The activation of STING leads to the production of potent immune-modulatory proteins, such as interferon (IFN), which can drive innate and adaptive immune responses, including antiviral and antitumor responses.  Hence, modulation of CDN pathways could be a basis of cancer therapy.  A potential problem with using CDN to bind to STING receptors and activate the cell's immune response is that the CDN doesn't last long, as it is hydrolyzed before it can bind to and act on tumor cells.  One way to stabilize them is to encapsulate them in nanovesicles, which persist longer and hence enhance their antitumor properties.  One such nanoparticle structure has shown great antitumor properties in mice.  The nanorod consists of negatively charged CDNs complexed with Mn²⁺ ions and 11-mer His peptides.  The resulting structures are encapsulated in a bilayer membrane, to which polyethylene glycol is attached to stabilize the structure.  Figure \(\PageIndex{2x}\) below shows an image made using Gemini based on the description of the nanorod in this research article:  Intermetallic nanoassemblies potentiate systemic STING activation.

Figure \(\PageIndex{2x}\):  xsxxxx.  Image made using Gemini based on the description of the nanodisc from the article: Intermetallic nanoassemblies potentiate systemic STING activation.

Now, viruses would deploy strategies to inhibit the host's innate immune responses.  One of these is the cleavage of the phosphodiester bonds in the small signaling molecules like cGAMP.  FoldSeek has identified LigT-like phosphodiesterases (PDEs) that cleave these small signaling molecules and prevent the activation of the host's innate immune system in the presence of viral DNA.  Figure \(\PageIndex{22}\) shows a phylogenetic tree of previously unknown LigT-like PDEs produced by many viruses.  They have activity against a wide range of cyclic nucleotides, including cGAMP.

Figure \(\PageIndex{22}\): A phylogenetic tree showing the polyphyletic lineages of LigT-like PDEs. Shaded boxes indicate viral taxa. The red residues in each protein structure are the conserved catalytic histidines. Units are substitutions per residue. The tree is colored according to bootstrap values. NCBI Protein accessions: YP_008798230, YP_002302228, YP_009021100, YP_003406995, NP_049750, YP_009047207, YP_009046269 and YP_009824980.  Nomburg, J. et al., ibid.

The LigT-like PDEs, first characterized in E. Coli, are named after the bacterial LigT protein, first characterized in E. coli. These enzymes have a common fold with two His-X-Ser/Thr motifs in the active site but don't have similar linear sequences. Hence, many have unknown functions (perhaps until this recent study)

Figure \(\PageIndex{23}\) shows an interactive iCn3D model of E.coli LigT complexed with 3'-AMP (5LDO). 3'-AMP is shown in spacefill.

Figure \(\PageIndex{23}\): E.coli LigT complexed with 3'-AMP (5LDO). (Copyright; author via source). Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/i...ACY6x3zhFdp6h7

Now compare this to the hypothetical LigT-PDE from pigeonpox virus (protein_HM89_gp030__YP_009046269__Pigeonpox_virus__10264) as shown in Figure \(\PageIndex{24}\) below.

Figure \(\PageIndex{24}\): Predicted structure of hypothetical LigT-PDE from pigeonpox virus.  

The PDB file of this pigeon pox protein (hypothetical_protein_HM89_gp030__YP_009046269__Pigeonpox_virus__10264) was downloaded along with all of the predicted structures from https://zenodo.org/records/10291581

To see an iCn3D model of the hypothetical LigT-PDE from the pigeonpox virus, follow these steps:

• open iCn3D
• Download this file to your computer's download folder.   IMPORTANT: If the file opens as an image in a new browser window, right-click the image and save the file to download it!
• File, Open File, iCn3D png (appendable), and choose the downloaded file

The E. Coli and pigeonpox virus proteins are similar in size and shape but have different secondary structures.  Note, however, the similarities and alignments between the active site histidines of both. 

 

---

 

TLRS and mRNA vaccines

Messenger RNA vaccines against the SARS-Cov2 spike protein have probably saved up to 20 million lives in the first year of the COVID-19 pandemic. The development of mRNA vaccines is a great scientific achievement that requires decades of fundamental and applied research by many scientists.

Vaccines usually are composed of target proteins from a virus, for example. Instead of delivering an actual protein, whose actual development and mass production takes years, why not use mRNA that encodes a viral protein or fragments of it? The idea has been around for a long time. The problem is that RNAs, with their 2'-OH on the ribose ring, are very labile and degrade easily. In addition, injecting RNA into a patient causes a significant immune response to the RNA. We mount immune responses to foreign viral RNA through our TLR receptors (TL3, 7, and 8), but what is needed is an immune response to the protein made from the injected mRNA, not to the RNA. Yet, we don't make an immune response against our RNA. Why?

Two major problems had to be solved (and a host of others as well) to make mRNA vaccines: the stability of the vaccines and our immune response to them. A hint comes from the observation that TLRs recognize non- or undermethylated DNA found in bacteria. Methylated CpG motifs in DNA do not stimulate an immune response. Katalin Karikó, Michael Buckstein, Houping Ni, and Drew Weissman reported in 2005 that the incorporation of methylated (m) and modified nucleosides m5C, m6A, m5U, s2U, and pseudouridine ablated the immune response to the RNA. This opened the door to mRNA vaccines. The paper was rejected by Nature and Science but published in Immunity (Vol. 23, 165–175, August 2005. DOI 10.1016/j.immuni.2005.06.008). The last line from the paper was truly prophetic: "Insights gained from this study could advance our understanding of autoimmune diseases where nucleic acids play a prominent role in the pathogenesis, determine a role for nucleoside modifications in viral RNA, and give future directions into the design of therapeutic RNAs".

Katalin Karikó and Drew Weissman were awarded the 2021 Lasker–DeBakey Clinical Medical Research Award and the Nobel Prize in Medicine in 2023 for their fundamental research that has saved many of us. See Chapter 9.1 for more details about the role of pseudouridine in mRNA vaccines.

Inflammasome

Think of what you would want your immune system to protect you from. Of course, there are pathogens like viruses, bacteria, and fungi. And, of course, you want to protect yourself from yourself, because you don't want to activate your immune system with self-antigens. But what about "non-biological" molecules like silica or asbestos, whose presence might be deleterious? What about normal biomolecules (proteins, nucleic acids) that suddenly find themselves in the wrong cellular location due to cell death by necrosis or physical injury?

In the previous section, we discussed how innate system immune cells (dendritic cells, macrophages, eosinophils, etc.) have receptors that recognize common pathogen-associated molecular patterns (PAMPs) such as lipopolysaccharides (LPS) on the surface of bacteria, mannose on bacteria and yeast, flagellin from bacterial flagella, dsRNA (from viruses) and nonmethylated CpG motifs in bacterial DNA. These antigens are recognized by pattern recognition receptors (PRRs) - specifically the Toll-like Receptors (TLRs) 1-10. These include plasma membrane TLRs (TL4 for LPS, TL5 for flagellin, TLR 1, 2, and 6 for membrane and wall components of fungi and bacteria) and intracellular endosomal TLRs (TLR3 for dsRNA, TLR 7 and 8 for ssRNA and TLR9 for dsDNA)

Damage-associated molecular patterns (DAMPs) are typically found on molecules released from cells or intracellular compartments following cellular damage (hence the name DAMP). Many are nuclear or cytoplasmic proteins released from the cells. These would now find themselves in a more oxidizing environment, further changing their properties. Common DAMP proteins include heat shock proteins, histones high mobility group proteins (both nuclear), and cytoskeletal proteins. What non-protein molecules might be released from damaged cells that pose problems? Here are some other common non-protein DAMPS: ATP, uric acid, heparin sulfate, DNA, and cholesterol crystals. In the wrong location, these can be considered danger signals.

If TLRs recognize PAMPs, what recognizes DAMPs? They are recognized by another type of intracellular pattern recognition receptor (PRR) called NOD (Nucleotide-binding Oligomerization Domain (NOD)-Like Receptors, or NLRs). NLRs also recognize PAMPs. The proteins are also named the Nucleotide-binding domain (NBD), and Leucine-Rich repeat (LRR)–containing proteins (NLRs). This family of proteins participates in the formation of a large protein structure called the inflammasome. (Sorry about the multiple abbreviations and naming systems!)

As both PAMPs and DAMPs pose dangers, it would make sense that, once they recognize their cognate PRRs (TLRs and NLRs, respectively), pathways emanating from the occupied receptors might converge in a common effector system that releases inflammatory cytokines from immune cells. Since uncontrolled immune effector release during an inflammatory response can be dangerous, requiring two signals to trigger cytokine release from the cell can sometimes be helpful. We've seen these two-signal requirements for T cell activation.

Two such inflammatory cytokines are Interleukin 1-beta (IL-1β) and IL-18. Activation of TLRs by a PAMP triggers activation of a potent immune cell transcription factor, NF-kbeta, which leads to transcription of the gene for the precursor of the cytokine, pro-interleukin 1-beta. Without a specific proteolytic cleavage, the active cytokine will not be released from the cell.

The protease required for this cleavage is activated by a signal arising when a DAMP activates a NLR, which then, through a sequence of interactions, leads to the proteolytic activation of another inactive protease, procaspase 1, on a large multi-protein complex called the inflammasome. (In later chapters, we will see other such protein complexes with targeted activities - including the spliceosome, which splices RNA to produce mRNA, and the proteasome, which conducts controlled intracellular proteolysis). The activated inflammasome activates procaspase, producing the active caspase (a cysteine-aspartic protease).

The convergence of the signals from the PAMP activation of a TLR and DAMP activation of a NLR at the inflammasome is shown in Figure \(\PageIndex{25}\).

Figure \(\PageIndex{25}\): Signals from PAMP activation of a TLR and DAMP activation of a NLR at the inflammasome

The active cytokine interleukin 1-beta helps recruit innate immune cells to the site of infection. It also affects the activity of immune cells in the adaptive immune response (T and B cells). Active IL-18 increases another cytokine, interferon-gamma, and enhances the activity of T cells that kill other cells.

The focus of this chapter is on binding interactions and their biological consequences. From that perspective, this section will address

• the structure and activity of caspases, which activate the pro-cytokine prointerleukin 1 beta,
• the structure and ligands for the NLRs, the structure and properties of the inflammasome, and finally
• how "danger" molecules such as ATP and crystals (cholesterol, silica) activate the inflammasome.

Unfortunately, there are many proteins involved with crazy acronyms for names. These proteins have multiple domains, and many of them have multiple names. Sorry in advance!

Caspases

Caspases (Cys-asp-proteases), not to be confused with Cas9 (CRISPR-associated protein 9, an RNA-guided DNA endonuclease), are a protease that, when active, can lead to cell death or, in a less austere outcome, initiate the inflammatory response (sometimes good, often bad, or even fatal). They have an active site nucleophilic Cys that cleaves peptide bonds after an Asp in target proteins. All caspases (13 in humans) have an N-terminal pro-domain followed by large and small protease catalytic domain subunits. As with other proteases, it is found as an inactive zymogen. Why is this important?

To become activated, they are recruited to a scaffolding protein, which is activated by the removal of the zymogen's N-terminal domain, followed by a second cut between the large and small catalytic subunits. The enzyme that does this is caspase itself in an autocatalytic step. There are three caspase types, two of which are involved in programmed cell death. We'll discuss the processing of Caspase-1 by inflammatory cytokines. Once activated, the initiator caspase activates other effector (executioner) caspases in the cell. The inflammasome activates caspase 1.

Two major domains are found in Caspase 1: the caspase recruitment domain (CARD), which mediates self-interaction with scaffold and adaptor proteins in the inflammasome for activation, and a proteolytic catalytic domain, as shown in Figure \(\PageIndex{26}\). All domain structures in the section were obtained using Conserved Domains from the NCBI (https://www.ncbi.nlm.nih.gov/Structure/cdd/wrpsb.cgi) or the Simple Modular Architecture Research Tool (SMART) at the EMBL ( http://smart.embl-heidelberg.de/smart/set_mode.cgi?NORMAL=1 ). Uniprot was used for protein (FASTA) sequences (http://www.uniprot.org/uniprot/). We will see the CARD domain often.

Figure \(\PageIndex{26}\): Domains structure of Caspase 1

NOD-like receptor proteins (NLRPs)

The NOD-like receptor proteins (NLRPs) are a family of proteins with similar domain structures. The structures and abbreviations used for the molecular players in inflammasome activation are complicated and confusing. Different programs show different domains, which adds to the complexity. We will attempt to reduce confusion by showing only domain structure diagrams, even if some diagrams show different domains for the same protein. Remember, domains are calculated from structure, so different algorithms using different databases might return different domain structures. Table \(\PageIndex{2}\) below shows the domain structure for NLRPs.

NLRP1,2,3
NLRP1 (NALP1)
NLRP3 (NALP3)
NLRP4
NAIP1
NAIP2

Table \(\PageIndex{2}\): Domain structures of the NLRPs

Another protein in the NLR family is NAIP (neuronal apoptosis inhibitor protein). The domain structure for NAIP1 is shown in Table \(\PageIndex{2}\) above. In contrast to the other NLRs for which specific ligands have not yet been found, several NAIPs have been shown to bind specific PAMPs. NAIP1 binds the needle protein CprI from C.violaceum, which starts to drive the assembly of the NLRC4 inflammasome. NAIP2 binds the inner rod protein of the bacterial type III secretion system (PrgJ in Salmonella typhimurium). NAIP5 and NAIP6 bind bacterial flagellin (which, for Salmonella typhimurium is the protein FliC). In the second domain representation, AAA is ATP-associated activities in the cell (otherwise denoted as the NACHT domain in the top representation).

NAIP2 interacts with another adapter NLR family protein, NLRC4 (NLR family CARD domain-containing protein), to form the inflammasome. The domain structure of NAIP2 is shown in Figure \(\PageIndex{x}\) below:

Note that many of these proteins share common domains:

• Pyrin-NALP - Pyrin domains on different proteins self-associate through inter-protein Pyrin:Pyrin interactions
• NACHT - This domain contains about 300-400 amino acids and can bind ATP and may cleave it (i.e., act as an ATPase)
• LRR - for Leucine Rich Repeat. These 20-30 amino acid repeats may occur up to 45 times in a given protein. They fold into an arc shape and seem to facilitate protein:protein interactions. On the concave side of the arc, they have a parallel beta sheet, while on the convex side, they have an alpha helix. They also appear to be involved in the binding of PAMPS and DAMPs;
• CARD - for caspase activation and recruitment. CARD domains on different proteins self-associate through inter-protein CARD:CARD interactions;
• BIR - Baculoviral inhibition of apoptosis protein repeat;
• ASC - Apoptosis-associated speck-like protein containing a CARD Adapter domain, allowing it to interact with other proteins with a CARD domain.

ASC Adaptor Protein

Small adapter proteins like ASC, which contain a CARD domain, mediate caspase binding in the apoptosome (involved in apoptosis or programmed cell death) and in the inflammasome. This smaller protein has two domains, a pyrin domain and a CARD domain, as shown in Figure \(\PageIndex{27}\). It is required for the recruitment of caspase-1 to some inflammasomes (for example, ones that contain NLRP2 and NLRP3

Figure \(\PageIndex{27}\): Domain structure of the small adapter proteins ASC

The Active Inflammasome

The active inflammasome generally consists of three different kinds of proteins, some present in multiple copies: NLRPs, adapter proteins like ASC, and procaspases. They may also contain additional recruitment and ligand sensor proteins. We'll discuss two types of inflammasome using different NLRPs, the NLRP4 and NLRP3 inflammasome.

NLRP4 Inflammasome

Some of the best structures (obtained by cryomicroscopy) are for the NAIP2:NLRP4 inflammasome. Figure \(\PageIndex{28}\) shows part of the complex consisting of 11 NLRP4 subunits arranged in a large ring. The actual biological complex has 1 NAIP2 subunit and 10 NLRP4s.

Figure \(\PageIndex{28}\): Part of the complex of the NAIP2:NLRP4 inflammasome showing 11 NLRP4 subunits arranged in a large ring

Figure \(\PageIndex{29}\) shows an interactive iCn3D model of the activated NAIP2-NLRC4 inflammasome (3JBL))

Figure \(\PageIndex{29}\): Activated NAIP2-NLRC4 inflammasome (3JBL). (Copyright; author via source). Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/i...8Ae7vY4aX1SLX8

How does this structure arise? Presumably, it would not exist without a PAMP or DAMP to minimize immune-mediated inflammatory damage. Data suggests that the bacterial protein PrgJ (denoted PrgX in the figure below) binds to its receptor, NAIP2, altering its conformation as shown in Figure \(\PageIndex{30}\). This binary complex presents an asymmetric electrostatic surface that allows a loose association with NLRP4, which, upon a conformational change, leads to a tighter binding interaction. Nine more NLRP4s bind similarly to form the 11-subunit ring structure.

Figure \(\PageIndex{30}\): Formation of the NAIP2:NLRP4 ring

The complementary electrostatic interactions between two of the many NLRC4 subunit monomers in the NLRP4 inflammasome are depicted in Figure \(\PageIndex{31}\).

Figure \(\PageIndex{31}\): Complementary electrostatic interactions between two of the many NLRC4 subunit monomers in the NLRP4 inflammasome

The top right panel shows two of the NLRC 4monomers (of the 12 in the Jsmol model) bound to each other, with the A subunit's concave inner face interacting with the B subunit's convex outer face. Only the interactions at the top of the dimers are highlighted for simplicity. The other panels show each monomer's electrostatic potential surface (red indicating negative and blue positive) on a sliding scale of -5 to +5 (images created using the PDB2PQR Server and Pymol).

The depicted negative (red) electrostatic potential outer surface on one NLRC4 monomer, complementary to the positive (inner) surface on the other NLRC4 subunit, is outlined in each panel in red or blue ellipses, respectively. The curved tertiary structure of the proteins and the opposing electrostatic surface potentials of opposite faces commit the subunits to form a large ringed 12-mer inflammasome core.

Note that this assembled ring brings together the CARD (caspase recruitment domain, yellow circles/spheres), which can interact with the CARD domain of the procaspase protein through CARD:CARD inter-protein interactions (think of a stack of playing cards all stuck together in a deck of cards) as shown in Figure \(\PageIndex{32}\).

Figure \(\PageIndex{32}\): Interactions of procaspase with the NAIP2:NLRP4 ring

Once assembled, proximal procaspases autocatalytically convert procaspase 1 into active caspase 1, which can, by proteolysis, activate the cytokines interleukin 1 beta and interleukin 18, releasing them from the cell. Remember that procytokines are produced only when their genes are transcribed following activation of the transcription factor NF-kappa beta by PAMP binding to a TLR.

NLRP3 Inflammasomes

Figure \(\PageIndex{33}\) shows an interactive iCn3D model of the NLRP3 double-ring cage, 6-fold (12-mer) (7LFH)

Figure \(\PageIndex{33}\): NLRP3 double-ring cage, 6-fold (12-mer) (7LFH). (Copyright; author via source). Click the image for a popup or use this external link: https://structure.ncbi.nlm.nih.gov/icn3d/share.html?DEbdkUoBtqRQ9bu59

The full-length mouse NLRP3 consists of 12- to 16-mer organized in a double-ring cage. It is held together by interactions between the leucine-rich repeats (LRR) domains. The structure shields the pyrin domains, so they will not be activated without appropriate signals. The complex is also localized to the membrane.

NLRP4 inflammasomes require specific PAMPs/DAMPs, such as bacterial flagellin, for activation.  In contrast, NLRP3 inflammasomes seem to be activated by multiple signals and cellular stress.  These include extracellular ATP, particulates like silica and uric acid crystals, toxins, and intracellular mediators of distress, often caused by pathogens. These include reactive oxygen species, organelle damage, etc. Given the large number of microbes that activate the NLRP3 inflammasome, it has been suggested that the signal that triggers NLRP3 is indirect.   One such indirect signal is the level of K⁺ ions in cells.

In normal cells, K⁺ ions are higher in the cytoplasm than outside the cell. Decreased potassium ion levels in cells due to efflux can activate NLRP3 inflammasomes. Other conditions include lysosomal rupture (perhaps associated with the cellular uptake of particles such as silica, uric acid, cholesterol crystals, and other "nanoparticles"), altered mitochondrial metabolism (which can lead to reactive oxygen species within the cell), etc. None of these danger triggers bind to NLPR3, but they somehow lead to downstream activation of NLPR3. NLRP3, hence, probably functions as a general sensor of cellular stress.

Inappropriate and chronic activation of inflammation has been associated with many diseases, such as cancer, cardiovascular disease, diabetes, and autoimmune diseases. Given the multiple types of signals that can activate the NLRP3 inflammasome, this complex is a focus of active drug development to identify inhibitors that can stop undesired inflammation. These inflammasomes are found in granulocytes, monocytes (macrophages), megakaryocytes, and dendritic cells.

Activated NLRP3 recruits the ASC Adaptor Protein, which in turn recruits and activates procaspase 1. NLRP3 has a pyrin, NACHT, and LRR domain. ASC has a pyrin and CARD domain. Active LRP3 can then recruit ASC via pyrin:pyrin inter-protein domain interactions. This allows the CARD domain of bound ASC to recruit procaspase via CARD:CARD interactions (remember that procaspase also has a CARD domain), forming the active NLRP3 inflammasome. An additional feature of NLRP3 inflammasome activation occurs when the transcription factor NFkB, activated by PAMPs (signal 1), induces transcription of both procytokines (IL-1β and IL-18) and NLRP3 itself.

Hence, two signals are again needed:

Signal 1

The first signals are the bacterial and viral (influenza virus, poliovirus, enterovirus, rhinovirus, human respiratory syncytial virus, etc.) PAMPs, which bind to TLRs and lead to the activation of the NFkb transcription factor. This activates not only the transcription of pro-interleukin 1-beat and interleukin 18 but also the transcription of the NLRP3 sensor itself.

Signal 2

Signal 2 is delivered to the sensor NLRP3 by PAMPs and DAMPs, indirectly. This leads to the assembly of the inflammasome. These DAMPs prime the activation of NLRP3 protein and subsequent formation of the active NLRP3 inflammasome. But what activates NLR3P3? After many studies, it became clear that the typical bacterial ligands that would activate TLRs and perhaps NLRs only prime NLRP3 for activation. They don't bind to it directly.

Extracellular ATP is a major activator of NLRP3. Nanoparticles are known to release ATP as well. Most studies show that K⁺ efflux from the cell is an early signal and that the NEK7, which phosphorylates other proteins, binds to NLRP3 after potassium ion efflux and activates it. Removing NEK7 stopped NLRP3 but not NLRP4 inflammasome activation. Although NLRP3 bound to NEK7 through the NEK7 catalytic domain, the activity of the catalytic domain of NEK7 was not needed.

What leads to K⁺ efflux? Let's back up to possible upstream events that could lead to efflux and try to find a link to ATP. The background for some of this material will be explored in future chapters. The following steps occur as shown in the figure and information below:

- solids such as silica, cholesterol crystals, uric acid crystals, and even aggregated proteins such as prions can be engulfed by monocytes/macrophages (much as they engulf bacteria as part of their immune function) in phagocytosis. The particles are enveloped in a plasma bilayer-derived membrane that buds off into the cell. This vesicle merges with a lysosome, which gets damaged in the process. They then release ATP into the cytoplasm;

- cytoplasmic ATP can then move outside of the cell through the glycoprotein membrane channel called pannexin 1;

- extracellular ATP can bind to another membrane protein called the P2X7 purinoceptor. This protein becomes a cation channel, allowing K⁺ efflux since the ion has a higher concentration inside the cell than outside, as illustrated in Figure \(\PageIndex{34}\). The extracellular ATP "gates" open the P2X7 cation channel. The pore-forming toxin nigericin from Streptomyces hygroscopicus also leads to potassium ion efflux. Likewise, pore-forming proteins from S. aureus (hemolysins) lead to potassium ion efflux and activation of the NLRP3 inflammasome. We will discuss membrane proteins in great detail in a later chapter.

Figure \(\PageIndex{34}\):Extracellular ATP triggers K⁺ efflux from cell

Other signals also activate the NLRP3 inflammasome. These include mitochondrial damage and the release of reactive oxygen species.

Summary

(Summary written by Claude, Sonnet 4.6, Anthropic)

This chapter examines the molecular mechanisms by which the innate immune system detects and responds to both foreign pathogens and self-derived danger signals — emphasizing the structural basis of pattern recognition, the logic of two-signal safety mechanisms, and the assembly of the inflammasome as a prototypical multi-protein signaling machine.

Pattern recognition in the innate immune system operates fundamentally differently from the adaptive immune response. Rather than generating tailored receptors for individual antigens through VDJ recombination, innate immune cells express a fixed repertoire of pattern recognition receptors (PRRs) that detect conserved molecular signatures shared across broad classes of pathogens. Toll-like receptors (TLRs 1–10) recognize pathogen-associated molecular patterns (PAMPs) at the cell surface and in intracellular endosomes. PAMPs detected include lipopolysaccharide (LPS) from Gram-negative bacteria (TLR4), flagellin (TLR5), bacterial and fungal membrane components (TLR1/2/6), viral dsRNA (TLR3), viral ssRNA (TLR7/8), and unmethylated CpG motifs in bacterial and viral DNA (TLR9). The structural basis of TLR3 recognition of dsRNA is particularly elegant: the ectodomain folds into a horseshoe-shaped solenoid of 23 leucine-rich repeats (LRRs) that dimerizes around the dsRNA ligand when it is at least 40–50 bp in length, with extensive N-linked glycosylation protecting most surfaces and leaving a carbohydrate-free face for dimerization. The recognition of dsRNA by TLR3 reflects the evolutionary logic of targeting molecular signatures absent from healthy host cells: most host RNA is single-stranded, so dsRNA (a byproduct of viral replication) serves as an unambiguous danger signal.

The discovery that modified nucleosides ablate TLR-mediated immune recognition of RNA — published by Karikó, Buckstein, Ni, and Weissman in 2005 — resolved a critical barrier to mRNA vaccine development. Just as bacteria methylate their DNA to prevent their own restriction enzymes from cleaving it, host cells modify their RNA with pseudouridine, m5C, m6A, and other modifications that prevent recognition by TLR3/7/8. Incorporation of pseudouridine into synthetic mRNA prevents innate immune recognition while preserving translational efficiency — the molecular innovation that enabled the mRNA vaccines that saved millions of lives during the COVID-19 pandemic and earned Karikó and Weissman the 2023 Nobel Prize in Physiology or Medicine.

A parallel innate surveillance axis detects cytoplasmic viral nucleic acids. When viral dsDNA enters the cytoplasm, it binds and activates cGAS (cyclic GMP-AMP synthase), which dimerizes and produces the cyclic dinucleotide second messenger 2',3'-cGAMP. cGAMP then binds the ER/mitochondria-associated effector protein STING (stimulator of interferon genes), triggering interferon production and antiviral gene expression. Similarly, viral dsRNA activates OAS (oligoadenylate synthase) to produce 2',5'-oligoadenylate, which activates the endoribonuclease RNaseL to degrade viral and cellular RNA. Prokaryotes use analogous cyclic dinucleotide signaling (3',3'-cGAMP synthesized by CD-NTases) as part of their anti-phage defense. Viruses counter these pathways by deploying LigT-like phosphodiesterases that hydrolyze cGAMP and related signaling molecules, preventing innate immune activation — an evolutionarily ancient viral escape mechanism discovered by structural comparison using FoldSeek across newly predicted viral protein structures, many of which share the conserved His-X-Ser/Thr active site motif despite unrelated sequences.

The inflammasome is a large multi-protein platform that integrates two distinct classes of danger signals — PAMPs (from pathogens, recognized by TLRs and NLRs) and DAMPs (from self-derived damage signals, recognized primarily by NLRs) — to trigger the proteolytic release of potent pro-inflammatory cytokines. The requirement for two converging signals before cytokine release provides a critical safety mechanism against inadvertent activation: Signal 1 activates the transcription factor NF-κB through TLR:PAMP binding, driving transcription of the cytokine precursors pro-IL-1β and pro-IL-18 as well as NLRP3 itself; Signal 2 comes from DAMPs or indirect stress signals that activate NLRP3 (or other NLRs), triggering inflammasome assembly.

Two distinct inflammasome architectures illustrate how domain-mediated protein-protein interactions drive the assembly of these signaling machines. The NAIP2:NLRC4 inflammasome responds to specific bacterial PAMPs: PrgJ (bacterial rod protein) or flagellin binds NAIP2, inducing a conformational change that creates a complementary electrostatic interface (a positively charged concave face on one NLRC4 subunit contacting a negatively charged convex face on the next), recruiting ten additional NLRC4 subunits sequentially to form a precise 11-mer ring structure. The assembled ring clusters the CARD (caspase activation and recruitment) domains of NLRC4, which recruit procaspase-1 through CARD:CARD interdomain interactions. The NLRP3 inflammasome is a broader sensor of cellular stress, responding to extracellular ATP, uric acid crystals, cholesterol crystals, silica particles, lysosomal rupture, and reactive oxygen species — none of which binds NLRP3 directly. Instead, these stressors converge on K⁺ efflux as the common intracellular signal: DAMPs trigger lysosomal rupture and ATP release into the cytoplasm; cytoplasmic ATP exits through the pannexin-1 channel; extracellular ATP gates open the P2X7 purinoceptor cation channel, driving K⁺ out of the cell; low intracellular K⁺ promotes binding of the kinase NEK7 to NLRP3, activating it. Activated NLRP3 then recruits the adaptor protein ASC through pyrin:pyrin domain interactions; ASC's CARD domain recruits procaspase-1 through CARD:CARD interactions, assembling the active NLRP3 inflammasome. In both inflammasome types, proximity-induced autocatalytic activation converts procaspase-1 to active caspase-1, which proteolytically cleaves pro-IL-1β and pro-IL-18 into active cytokines that are released from the cell to recruit and activate additional immune cells. Chronic NLRP3 inflammasome activation has been implicated in cancer, cardiovascular disease, type 2 diabetes, and autoimmune diseases, making it a major target for anti-inflammatory drug development.