CBASS to cGAS-STING: The Origins and Mechanisms of Nucleotide Second Messenger Immune Signaling

Abstract

Host defense against viral pathogens is an essential function for all living organisms. In cell-intrinsic innate immunity, dedicated sensor proteins recognize molecular signatures of infection and communicate to downstream adaptor or effector proteins to activate immune defense. Remarkably, recent evidence demonstrates that much of the core machinery of innate immunity is shared across eukaryotic and prokaryotic domains of life. Here, we review a pioneering example of evolutionary conservation in innate immunity: the animal cGAS-STING (cyclic GMP-AMP synthase-stimulator of interferon genes) signaling pathway and its ancestor in bacteria, CBASS (cyclic nucleotide-based antiphage signaling system) antiphage defense. We discuss the unique mechanism by which animal cGLRs (cGAS-like receptors) and bacterial CD-NTases (cGAS/dinucleotide-cyclase in Vibrio (DncV)-like nucleotidyltransferases) in these pathways link pathogen detection with immune activation using nucleotide second messenger signals. Comparing the biochemical, structural, and mechanistic details of cGAS-STING, cGLR signaling, and CBASS, we highlight emerging questions in the field and examine evolutionary pressures that may have shaped the origins of nucleotide second messenger signaling in antiviral defense.

Keywords: CBASS; STING; antiviral defense; cGAS; cGLR; innate immunity; nucleotide second messenger.

Fig. 1. Nucleotide second messenger signaling in cGLR-STING and CBASS immunity.

In animals, cGAS and cGLR pattern-recognition receptors detect molecular signatures of infection and synthesize cyclic dinucleotide second messengers. Upon binding nucleotide second messenger signals the adaptor protein STING activates antiviral transcription programs through NF-κB and IRF3. cGLR enzymes and STING are encoded at separate genomic loci: in humans cGAS is encoded on chromosome 6 (C6) and STING is encoded on chromosome 5 (C5). In bacterial CBASS systems, CD-NTase enzymes are activated by phage infection to synthesize cyclic di- or trinucleotide second messengers that are specifically recognized by adjacently encoded Cap (CD-NTase-associated protein) effectors. CBASS effectors target essential cellular components to induce abortive infection and protect bacterial populations from phage infection. CBASS operons universally encode a CD-NTase and an effector protein and are classified by regulatory Cap proteins. Type I systems contain no additional proteins. Type II systems encode Cap2, homologous to eukaryotic ubiquitin-activating (E1) and ubiquitin-conjugating (E2) enzymes, and Cap3, homologous to the eukaryotic JAB deubiquitinating endopeptidase. Type III systems encode Cap6, homologous to eukaryotic Trip13 AAA+ ATPases, and Cap7, homologous to eukaryotic HORMA domains. Some Type III systems additionally encode Cap8, a second divergent HORMA-like protein. Type IV systems encode proteins with predicted roles in converting and exchanging modified nucleobases: Cap9, a protein with a QueC enzymatic domain, and Cap10, a protein with a TGT enzymatic domain. Some Type IV systems additionally encode Cap10, a predicted N-glycosylase / DNA lyase. Cell-intrinsic immune pathways must connect pathogen sensing with a downstream immune response. Many classical pathways rely on direct protein–protein interactions between sensor proteins and adaptor proteins that activate transcription factors to induce antiviral gene programs (eukaryotes), or effector proteins that directly target invading viruses (prokaryotes). Animal cGLR-STING and bacterial CBASS pathways subvert this paradigm: sensor cGLRs or CD-NTases communicate immune activation to downstream adaptor or effector proteins using nucleotide second messenger signals. Checked boxes indicate the evolutionary advantages to this form of immune signaling in either eukaryotes or prokaryotes, discussed throughout this review. Abbreviations: cGLR, cGAS-like receptor; STING, Stimulator-of-Interferon-Genes; CBASS, cyclic nucleotide-based antiphage signaling system; cGAS, cyclic GMP-AMP synthase; NF-κB, nuclear factor κ-light-chain-enhancer of activated B cells; IRF3, interferon regulatory factor 3; CD-NTase, cGAS/DncV-like nucleotidyltransferase; JAB, Janus tyrosine kinase binding protein; HORMA, HOp1p, Rev7p and MAd2 protein; AAA+, ATPases Associated with diverse cellular Activities; TGT, tRNA-Guanine Transglycosylase.

Fig. 2. Pathogen sensing and regulation of cGLR and CD-NTase activity.

In animals, cGAS and cGLRs directly detect molecular signatures of pathogen infection, including dsDNA and dsRNA. In bacteria, the molecular mechanism by which CD-NTases detect phage infection is currently unclear. Recent evidence supports that a CD-NTase enzyme from Clade E (CdnE) directly detects the phage-derived cabRNA. Inset box: Pathogen sensing by cGLRs and CD-NTases evolves independently of downstream signaling components. Crystal structures of the human cGAS–DNA complex (PDB: 6CTA (44)) and the CD-NTase DncV (PDBs: 4TY0 / 4U0M (45, 46)) reveal close structural homology between animal cGLR enzymes and bacterial CD-NTases. A “spine” helix bridges an N-terminal nucleotidyltransferase core and a C-terminal helix bundle, enclosing a catalytic pocket that supports nucleotide second messenger synthesis. In cGAS, a Zn-ribbon motif inserts into the ligand binding groove to control dsDNA recognition. β-sheets highlighted in yellow; Zn-ribbon highlighted in magenta. Inset box: cGLRs and CD-NTase enzymes are ancestrally related to DNA-polβ family enzymes that carry out fundamental nucleotidyltransferase biochemistry, including DNA replication and RNA repair. Recognition of long dsDNAs by cGAS dimers controls enzyme oligomerization. The cGAS unstructured N-terminal tail (not shown) further drives the formation of molecular condensates, which promote robust enzymatic activity. cGAS binding partners including Ku proteins further promote cGAS-DNA binding and condensate formation. Inhibitory PTMs downregulate cGAS enzymatic activity and direct interactions with nucleosomes lock cGAS into a catalytically inactive conformation. In some CBASS systems, accessory Cap proteins regulate CD-NTase enzymatic activity. In Type II systems, Cap2 proteins prime nucleotide second messenger synthesis by conjugating the CD-NTase to an unknown protein substrate; Cap3 reverses this conjugation to downregulate activity. In Type III systems, peptide recognition by Cap6 promotes Cap6–CD-NTase complex formation to activate nucleotide second messenger synthesis; Cap7 disassembles the Cap6-CD-NTase complex. Abbreviations: cGAS, cyclic GMP-AMP synthase; cGLR, cGAS-like receptor; ds, double-stranded; PAMPs, pathogen-associated molecular patterns; CD-NTase, cGAS/DncV-like nucleotidyltransferase; CdnE, CD-NTase Clade E; cabRNA, CBASS-activating bacteriophage RNA; DncV, dinucleotide-cyclase in Vibrio; N-terminal, amino-terminal; C-terminal, carboxy-terminal; Zn, zinc; PTMs, post-translational modifications; Cap, CD-NTase-associated protein; Gl, glycan; Ub, ubiquitin; P, phosphate; Ac, acetyl-group.

Fig. 3. Nucleotide second messenger signals in cGLR and CBASS immunity.

cGAS, cGLR, and CD-NTase enzymes synthesize nucleotide second messenger signals in a conserved multi-step reaction. Inset box: cGLR and CD-NTase immune signals are synthesized from common ribonucleotide substrates. A catalytic triad of aspartate (D) and glutamate (E) residues coordinate either magnesium or manganese catalytic ions to sequentially synthesize 2′–5′ or 3′–5′ phosphodiester linkages between nucleotide substrates. Inset right: the human cGAS active site coordinating ATP and GTP substrates to synthesize first pppG[2′–5′]pA bond. Inset left: the DncV active site coordinating substrates to synthesize the first pppA[3′–5′]pG bond. Inset box: the multi-turnover synthesis of nucleotide second messengers dramatically amplifies immune signaling. Confirmed animal cGLR signals are shown with corresponding source organisms: 2′3′-cGAMP, diverse animals including humans, vertebrate fish, beetles, and Nematostella; 2′3′-cUA, C. gigas and C. virginica; 3′3′-cUA, S. pistallata; 3′2′-cGAMP, Drosophila. Confirmed bacterial CD-NTase signals are shown, including diverse cyclic di- and trinucleotides. Inset box: cGLR and CD-NTase signals rapidly evolve, likely to control divergent immune pathways and evade viral antagonism. Both animal viruses and bacteriophages encode nucleases that cleave cGLR and CD-NTase nucleotide second messengers to antagonize immune signaling. In animals, poxin efficiently cleaves 2′3′-cGAMP but not 3′2′-cGAMP. In bacteria, Acb1 efficiently cleaves nucleotide second messengers that incorporate an adenosine base but not 3′3′-cGG or 3′3′-cUU. In animals, the cGAS product 2′3′-cGAMP plays an important role as an intercellular immune signal. 2′3′-cGAMP is exported and imported through a number of channels, incorporated into viral particles, and transits to neighboring cells through gap junctions. The host-encoded nuclease ENPP1 cleaves 2′3′-cGAMP to regulate the available extracellular pool of this immune signal. Inset box: intercellular signaling is an evolutionary advantage of nucleotide second messenger immune signals demonstrated in animals but not in bacteria. Abbreviations: cGAS, cyclic GMP-AMP synthase; cGLR, cGAS-like receptor; CD-NTase, cGAS/DncV-like nucleotidyltransferase; DncV, dinucleotide-cyclase in Vibrio; cGAMP, cyclic GMP-AMP; cUA, cyclic UMP-AMP; cUU, cyclic UMP-UMP; cGG, clyclic GMP-GMP; cAAA, cyclic AMP-AMP-AMP; cAAG, cyclic AMP-AMP-GMP; poxin, poxvirus immune nuclease; Acb1, anti-CBASS protein 1; ABCC1, ATP-binding Cassette subfamily C member 1; ENPP1, ectonucleotide pyrophosphatase phosphodiesterase 1; SLC46A2, Solute Carrier family 46 member 2; SLC19A1, Solute Carrier family 19 member 1; LRRC8, Leucine-Rich-Repeat containing protein family 8.

Fig. 4. Receiving the signal: detection of nucleotide messengers by downstream receptors.

Specific domains in animal STING and bacterial CBASS receptor proteins control nucleotide second messenger sensing and downstream signaling or effector functions. In animal STING proteins the central ligand binding domain controls nucleotide second messenger sensing and N-terminal TM domains control localization to the ER membrane. In vertebrate STING proteins an unstructured C-terminal tail (CTT) recruits kinases to control downstream immune signaling. TIR domains in some invertebrate STING proteins have an uncharacterized immune mechanism. In bacteria, CBASS receptors are comprised of modular configurations of domains that sense nucleotide second messengers and domains with cell-death effector functions. TIR-STING, Cap12; TM-STING, Cap13; “nuclease”: REase-SAVED, Cap4; HNH-SAVED, Cap5. TM-SAVED, Cap14; TM-β-barrel, Cap15. 4 TM protein refers to S4TM; 2 TM protein refers to SLATT4; 3 TM protein refers to SLATT6. Diverse protein scaffolds in animals and bacteria specifically recognize cGLR and CBASS nucleotide second messenger signals. Inset box: nucleotide second messenger receptors are able to evolve independently of upstream pathogen sensing enzymes. Crystal structures of human STING in complex with 2′3′-cGAMP (PDB: 4KSY, (7)) and mouse RECON in complex with cAAG (PDB: 6M7K, (12)) demonstrate how nucleotide second messenger sensing domains in animals recognize either endogenously produced immune signals or bacterially-produced cross-kingdom signals. Structurally characterized nucleotide second messenger receptors in bacteria include STING in complex with 3′3′-c-di-GMP (PDB: 6WT4, (14)), SAVED in complex with 2′3′3′-cAAA (PDB: 6VM6, (17)). See also SAVED in complex with 3′2′-cGAMP (91)), and NucC in complex with 3′3′3′-cAAA (PDB: 6Q1H, (88)). Receptor protein oligomerization upon the sensing of nucleotide second messenger ligands is a conserved feature of animal cGLR-STING and bacterial CBASS signaling. In animal STING and bacterial TIR-STING (Cap12), conformational changes upon direct binding to nucleotide second messenger ligands induces head-to-head interactions between STING dimers and drives the formation of filaments. Vertebrate animal STING filament formation leads to the clustering of the CTT and the recruitment of the immune signaling kinase TBK1. Bacterial TIR-STING filament formation promotes cross-dimer TIR contacts to activate TIR NADase activity. Abbreviations: STING, Stimulator-of-Interferon-Genes; CBASS, cyclic nucleotide-based antiphage signaling system; N-terminal, amino-terminal; TM, transmembrane; C-terminal, carboxy-terminal; ER, endoplasmic reticulum; TIR, toll-interleukin-1 receptor; Cap, CD-NTase-associated protein; REase, restriction endonuclease; SAVED, synthase-associated and fused to various effector domains; HNH, nuclease with histidine (H) and asparagine (N) catalytic residues; NUDIX, [hydrolase that cleaves] nucleoside diphosphates linked to X; SLATT, synthase and LOG-Smf/DprA-associating two TM; cGLR, cGAS-like receptor; RECON, reductase controlling NF-κB; cGAMP, cyclic GMP-AMP; cAAG, cyclic AMP-AMP-GMP; NucC, nuclease, CD-NTase associated; TBK1, TANK-binding kinase 1; NADase, nicotinamide adenine dinucleotide hydrolase.

Fig. 5. Mechanisms of antiviral defense in cGAS and CBASS immunity.

Animal cGLR-STING signaling protects cells from infection through the induction of antiviral gene programs. The vertebrate STING CTT functions as a signaling platform, with distinct motifs to recruit the kinase TBK1 and the transcription factor IRF3. The STING CTT in ray-finned fish further has a motif to recruit TRAF6. Invertebrate STING proteins do not have a CTT. TBK1 dimers are recruited to the CTT following STING 2′3′-cGAMP binding and activate through a trans-autophosphorylation event. TBK1 then phosphorylates the upstream IRF3 motif to license the recruitment of IRF3, which is in turn phosphorylated and activated by TBK1. STING nucleotide second messenger binding further activates the transcription factors STAT6 and NF-κB through an unknown mechanism. Active STAT6, NF-κB, and IRF3 dimers translocate to the nucleus and induce the transcription of cytokines and type I interferons, resulting in inflammation, the induction of antiviral ISGs, and the stimulation of adaptive immunity. Following the activation of transcription factors at the Golgi membrane, STING traffics to the lysosome for degradation. Bacterial CBASS immunity relies on the degradation of essential cellular components by effector proteins. CBASS effectors include endonucleases, such as NucC and Cap4, that promiscuously degrade dsDNA; enzymes that degrade important cellular metabolites, such as Cap12 proteins that cleave NAD⁺; and proteins that disrupt inner membrane integrity either through TM domains, as does Cap15, or through phospholipase activity, as does CapV. The degradation of essential cellular components leads to cell death or dormancy, thereby limiting the production of progeny phage particles and halting viral spread throughout the bacterial population. Abbreviations: cGLR, cGAS-like receptor; STING, Stimulator-of-Interferon-Genes; CTT, C-terminal tail; TBK1, TANK-binding kinase 1; TRAF6, TNF-receptor associated factor 6; cGAMP, cyclic GMP-AMP; IRF3, interferon regulatory factor 3; STAT6, signal transducer and activator of transcription 6; NF-κB, nuclear factor κ-light-chain-enhancer of activated B cells; ISG, interferon-stimulated gene; IFN, interferon; JAK, Janus kinase; STAT, signal transducer and activator or transcription; CBASS, cyclic nucleotide-based antiphage signaling system; NucC, nuclease, CD-NTase associated; Cap, CD-NTase-associated protein; ds, double-stranded; NAD+, nicotine adenine dinucleotide; TM, transmembrane.

Fig. 6. Evolution of CBASS and cGLR immune systems across domains of life.

In bacteria, CBASS systems are encoded in defense islands on host chromosomes or mobile genetic elements (MGEs) such as prophage, transposons, and plasmids. Defense islands move through bacterial populations by vertical and horizontal gene transfer with a gene composition rapidly evolving through genetic exchange. Data in the field support two evolutionary scenarios for the acquisition of cGLR and STING genes by the metazoan lineage from ancestral CBASS systems: 1. cGLR and STING genes were acquired by a progenitor archaeon from a CBASS system in alpha-proteobacteria during endosymbiotic events that founded the eukaryotic lineage. As eukaryotes evolved, cGLRs and STING were retained by metazoans and lost in plants, fungi, and protists. 2. cGLR and STING genes were acquired by an ancestral metazoan through a rare event of horizontal gene transfer from bacteria. A further open evolutionary question is whether cGLRs and STING were acquired jointly, from a single CBASS operon, or individually, from distinct operons. cGLRs have evolved throughout animals to form a diverse family of pattern recognition receptors, and STING proteins have evolved to control different mechanisms of antiviral immunity. Abbreviations: CBASS, cyclic nucleotide-based antiphage signaling system; cGLR, cGAS-like receptor; STING, Stimulator-of-Interferon-Genes