Discovery of phage determinants that confer sensitivity to bacterial immune systems.

Over the past few years, numerous anti-phage defense systems have been discovered in bacteria. Although the mechanism of defense for some of these systems is understood, a major unanswered question is how these systems sense phage infection. To systematically address this question, we isolated 177 phage mutants that escape 15 different defense systems. In many cases, these escaper phages were mutated in the gene sensed by the defense system, enabling us to map the phage determinants that confer sensitivity to bacterial immunity. Our data identify specificity determinants of diverse retron systems and reveal phage-encoded triggers for multiple abortive infection systems. We find general themes in phage sensing and demonstrate that mechanistically diverse systems have converged to sense either the core replication machinery of the phage, phage structural components, or host takeover mechanisms. Combining our data with previous findings, we formulate key principles on how bacterial immune systems sense phage invaders.

Cryo-EM structure of the RADAR supramolecular anti-phage defense complex.

RADAR is a two-protein bacterial defense system that was reported to defend against phage by "editing" messenger RNA. Here, we determine cryo-EM structures of the RADAR defense complex, revealing RdrA as a heptameric, two-layered AAA+ ATPase and RdrB as a dodecameric, hollow complex with twelve surface-exposed deaminase active sites. RdrA and RdrB join to form a giant assembly up to 10 MDa, with RdrA docked as a funnel over the RdrB active site. Surprisingly, our structures reveal an RdrB active site that targets mononucleotides. We show that RdrB catalyzes ATP-to-ITP conversion in vitro and induces the massive accumulation of inosine mononucleotides during phage infection in vivo, limiting phage replication. Our results define ATP mononucleotide deamination as a determinant of RADAR immunity and reveal supramolecular assembly of a nucleotide-modifying machine as a mechanism of anti-phage defense.

Cyclic CMP and cyclic UMP mediate bacterial immunity against phages.

The cyclic pyrimidines 3',5'-cyclic cytidine monophosphate (cCMP) and 3',5'-cyclic uridine monophosphate (cUMP) have been reported in multiple organisms and cell types. As opposed to the cyclic nucleotides 3',5'-cyclic adenosine monophosphate (cAMP) and 3',5'-cyclic guanosine monophosphate (cGMP), which are second messenger molecules with well-established regulatory roles across all domains of life, the biological role of cyclic pyrimidines has remained unclear. Here we report that cCMP and cUMP are second messengers functioning in bacterial immunity against viruses. We discovered a family of bacterial pyrimidine cyclase enzymes that specifically synthesize cCMP and cUMP following phage infection and demonstrate that these molecules activate immune effectors that execute an antiviral response. A crystal structure of a uridylate cyclase enzyme from this family explains the molecular mechanism of selectivity for pyrimidines as cyclization substrates. Defense systems encoding pyrimidine cyclases, denoted here Pycsar (pyrimidine cyclase system for antiphage resistance), are widespread in prokaryotes. Our results assign clear biological function to cCMP and cUMP as immunity signaling molecules in bacteria.

Structural basis of Gabija anti-phage defence and viral immune evasion.

Bacteria encode hundreds of diverse defence systems that protect them from viral infection and inhibit phage propagation1-5. Gabija is one of the most prevalent anti-phage defence systems, occurring in more than 15% of all sequenced bacterial and archaeal genomes1,6,7, but the molecular basis of how Gabija defends cells from viral infection remains poorly understood. Here we use X-ray crystallography and cryo-electron microscopy (cryo-EM) to define how Gabija proteins assemble into a supramolecular complex of around 500 kDa that degrades phage DNA. Gabija protein A (GajA) is a DNA endonuclease that tetramerizes to form the core of the anti-phage defence complex. Two sets of Gabija protein B (GajB) dimers dock at opposite sides of the complex and create a 4:4 GajA-GajB assembly (hereafter, GajAB) that is essential for phage resistance in vivo. We show that a phage-encoded protein, Gabija anti-defence 1 (Gad1), directly binds to the Gabija GajAB complex and inactivates defence. A cryo-EM structure of the virally inhibited state shows that Gad1 forms an octameric web that encases the GajAB complex and inhibits DNA recognition and cleavage. Our results reveal the structural basis of assembly of the Gabija anti-phage defence complex and define a unique mechanism of viral immune evasion.

Phages overcome bacterial immunity via diverse anti-defence proteins.

It was recently shown that bacteria use, apart from CRISPR-Cas and restriction systems, a considerable diversity of phage resistance systems1-4, but it is largely unknown how phages cope with this multilayered bacterial immunity. Here we analysed groups of closely related Bacillus phages that showed differential sensitivity to bacterial defence systems, and discovered four distinct families of anti-defence proteins that inhibit the Gabija, Thoeris and Hachiman systems. We show that these proteins Gad1, Gad2, Tad2 and Had1 efficiently cancel the defensive activity when co-expressed with the respective defence system or introduced into phage genomes. Homologues of these anti-defence proteins are found in hundreds of phages that infect taxonomically diverse bacterial species. We show that the anti-Gabija protein Gad1 blocks the ability of the Gabija defence complex to cleave phage-derived DNA. Our data further reveal that the anti-Thoeris protein Tad2 is a 'sponge' that sequesters the immune signalling molecules produced by Thoeris TIR-domain proteins in response to phage infection. Our results demonstrate that phages encode an arsenal of anti-defence proteins that can disable a variety of bacterial defence mechanisms.

Viruses inhibit TIR gcADPR signalling to overcome bacterial defence.

The Toll/interleukin-1 receptor (TIR) domain is a key component of immune receptors that identify pathogen invasion in bacteria, plants and animals1-3. In the bacterial antiphage system Thoeris, as well as in plants, recognition of infection stimulates TIR domains to produce an immune signalling molecule whose molecular structure remains elusive. This molecule binds and activates the Thoeris immune effector, which then executes the immune function1. We identified a large family of phage-encoded proteins, denoted here as Thoeris anti-defence 1 (Tad1), that inhibit Thoeris immunity. We found that Tad1 proteins are 'sponges' that bind and sequester the immune signalling molecule produced by TIR-domain proteins, thus decoupling phage sensing from immune effector activation and rendering Thoeris inactive. Tad1 can also efficiently sequester molecules derived from a plant TIR-domain protein, and a high-resolution crystal structure of Tad1 bound to a plant-derived molecule showed a unique chemical structure of 1 ''-2' glycocyclic ADPR (gcADPR). Our data furthermore suggest that Thoeris TIR proteins produce a closely related molecule, 1''-3' gcADPR, which activates ThsA an order of magnitude more efficiently than the plant-derived 1''-2' gcADPR. Our results define the chemical structure of a central immune signalling molecule and show a new mode of action by which pathogens can suppress host immunity.

Prokaryotic viperins produce diverse antiviral molecules.

Viperin is an interferon-induced cellular protein that is conserved in animals1. It has previously been shown to inhibit the replication of multiple viruses by producing the ribonucleotide 3'-deoxy-3',4'-didehydro (ddh)-cytidine triphosphate (ddhCTP), which acts as a chain terminator for viral RNA polymerase2. Here we show that eukaryotic viperin originated from a clade of bacterial and archaeal proteins that protect against phage infection. Prokaryotic viperins produce a set of modified ribonucleotides that include ddhCTP, ddh-guanosine triphosphate (ddhGTP) and ddh-uridine triphosphate (ddhUTP). We further show that prokaryotic viperins protect against T7 phage infection by inhibiting viral polymerase-dependent transcription, suggesting that it has an antiviral mechanism of action similar to that of animal viperin. Our results reveal a class of potential natural antiviral compounds produced by bacterial immune systems.

Antiviral activity of bacterial TIR domains via immune signalling molecules.

The Toll/interleukin-1 receptor (TIR) domain is a canonical component of animal and plant immune systems1,2. In plants, intracellular pathogen sensing by immune receptors triggers their TIR domains to generate a molecule that is a variant of cyclic ADP-ribose3,4. This molecule is hypothesized to mediate plant cell death through a pathway that has yet to be resolved5. TIR domains have also been shown to be involved in a bacterial anti-phage defence system called Thoeris6, but the mechanism of Thoeris defence remained unknown. Here we show that phage infection triggers Thoeris TIR-domain proteins to produce an isomer of cyclic ADP-ribose. This molecular signal activates a second protein, ThsA, which then depletes the cell of the essential molecule nicotinamide adenine dinucleotide (NAD) and leads to abortive infection and cell death. We also show that, similar to eukaryotic innate immune systems, bacterial TIR-domain proteins determine the immunological specificity to the invading pathogen. Our results describe an antiviral signalling pathway in bacteria, and suggest that the generation of intracellular signalling molecules is an ancient immunological function of TIR domains that is conserved in both plant and bacterial immunity.

Cyclic GMP-AMP signalling protects bacteria against viral infection.

The cyclic GMP-AMP synthase (cGAS)-STING pathway is a central component of the cell-autonomous innate immune system in animals1,2. The cGAS protein is a sensor of cytosolic viral DNA and, upon sensing DNA, it produces a cyclic GMP-AMP (cGAMP) signalling molecule that binds to the STING protein and activates the immune response3-5. The production of cGAMP has also been detected in bacteria6, and has been shown, in Vibrio cholerae, to activate a phospholipase that degrades the inner bacterial membrane7. However, the biological role of cGAMP signalling in bacteria remains unknown. Here we show that cGAMP signalling is part of an antiphage defence system that is common in bacteria. This system is composed of a four-gene operon that encodes the bacterial cGAS and the associated phospholipase, as well as two enzymes with the eukaryotic-like domains E1, E2 and JAB. We show that this operon confers resistance against a wide variety of phages. Phage infection triggers the production of cGAMP, which-in turn-activates the phospholipase, leading to a loss of membrane integrity and to cell death before completion of phage reproduction. Diverged versions of this system appear in more than 10% of prokaryotic genomes, and we show that variants with effectors other than phospholipase also protect against phage infection. Our results suggest that the eukaryotic cGAS-STING antiviral pathway has ancient evolutionary roots that stem from microbial defences against phages.

Communication between viruses guides lysis-lysogeny decisions.

Temperate viruses can become dormant in their host cells, a process called lysogeny. In every infection, such viruses decide between the lytic and the lysogenic cycles, that is, whether to replicate and lyse their host or to lysogenize and keep the host viable. Here we show that viruses (phages) of the SPbeta group use a small-molecule communication system to coordinate lysis-lysogeny decisions. During infection of its Bacillus host cell, the phage produces a six amino-acids-long communication peptide that is released into the medium. In subsequent infections, progeny phages measure the concentration of this peptide and lysogenize if the concentration is sufficiently high. We found that different phages encode different versions of the communication peptide, demonstrating a phage-specific peptide communication code for lysogeny decisions. We term this communication system the 'arbitrium' system, and further show that it is encoded by three phage genes: aimP, which produces the peptide; aimR, the intracellular peptide receptor; and aimX, a negative regulator of lysogeny. The arbitrium system enables a descendant phage to 'communicate' with its predecessors, that is, to estimate the amount of recent previous infections and hence decide whether to employ the lytic or lysogenic cycle.

CRISPR adaptation biases explain preference for acquisition of foreign DNA.

CRISPR-Cas (clustered, regularly interspaced short palindromic repeats coupled with CRISPR-associated proteins) is a bacterial immunity system that protects against invading phages or plasmids. In the process of CRISPR adaptation, short pieces of DNA ('spacers') are acquired from foreign elements and integrated into the CRISPR array. So far, it has remained a mystery how spacers are preferentially acquired from the foreign DNA while the self chromosome is avoided. Here we show that spacer acquisition is replication-dependent, and that DNA breaks formed at stalled replication forks promote spacer acquisition. Chromosomal hotspots of spacer acquisition were confined by Chi sites, which are sequence octamers highly enriched on the bacterial chromosome, suggesting that these sites limit spacer acquisition from self DNA. We further show that the avoidance of self is mediated by the RecBCD double-stranded DNA break repair complex. Our results suggest that, in Escherichia coli, acquisition of new spacers largely depends on RecBCD-mediated processing of double-stranded DNA breaks occurring primarily at replication forks, and that the preference for foreign DNA is achieved through the higher density of Chi sites on the self chromosome, in combination with the higher number of forks on the foreign DNA. This model explains the strong preference to acquire spacers both from high copy plasmids and from phages.

A vast collection of microbial genes that are toxic to bacteria.

In the process of clone-based genome sequencing, initial assemblies frequently contain cloning gaps that can be resolved using cloning-independent methods, but the reason for their occurrence is largely unknown. By analyzing 9,328,693 sequencing clones from 393 microbial genomes, we systematically mapped more than 15,000 genes residing in cloning gaps and experimentally showed that their expression products are toxic to the Escherichia coli host. A subset of these toxic sequences was further evaluated through a series of functional assays exploring the mechanisms of their toxicity. Among these genes, our assays revealed novel toxins and restriction enzymes, and new classes of small, non-coding toxic RNAs that reproducibly inhibit E. coli growth. Further analyses also revealed abundant, short, toxic DNA fragments that were predicted to suppress E. coli growth by interacting with the replication initiator DnaA. Our results show that cloning gaps, once considered the result of technical problems, actually serve as a rich source for the discovery of biotechnologically valuable functions, and suggest new modes of antimicrobial interventions.

The SARM1 TIR domain produces glycocyclic ADPR molecules as minor products.

Sterile alpha and TIR motif-containing 1 (SARM1) is a protein involved in programmed death of injured axons. Following axon injury or a drug-induced insult, the TIR domain of SARM1 degrades the essential molecule nicotinamide adenine dinucleotide (NAD+), leading to a form of axonal death called Wallerian degeneration. Degradation of NAD+ by SARM1 is essential for the Wallerian degeneration process, but accumulating evidence suggest that other activities of SARM1, beyond the mere degradation of NAD+, may be necessary for programmed axonal death. In this study we show that the TIR domains of both human and fruit fly SARM1 produce 1''-2' and 1''-3' glycocyclic ADP-ribose (gcADPR) molecules as minor products. As previously reported, we observed that SARM1 TIR domains mostly convert NAD+ to ADPR (for human SARM1) or cADPR (in the case of SARM1 from Drosophila melanogaster). However, we now show that human and Drosophila SARM1 additionally convert ~0.1-0.5% of NAD+ into gcADPR molecules. We find that SARM1 TIR domains produce gcADPR molecules both when purified in vitro and when expressed in bacterial cells. Given that gcADPR is a second messenger involved in programmed cell death in bacteria and likely in plants, we propose that gcADPR may play a role in SARM1-induced programmed axonal death in animals.

Self-targeting by CRISPR: gene regulation or autoimmunity?

The recently discovered prokaryotic immune system known as CRISPR (clustered regularly interspaced short palindromic repeats) is based on small RNAs ('spacers') that restrict phage and plasmid infection. It has been hypothesized that CRISPRs can also regulate self gene expression by utilizing spacers that target self genes. By analyzing CRISPRs from 330 organisms we found that one in every 250 spacers is self-targeting, and that such self-targeting occurs in 18% of all CRISPR-bearing organisms. However, complete lack of conservation across species, combined with abundance of degraded repeats near self-targeting spacers, suggests that self-targeting is a form of autoimmunity rather than a regulatory mechanism. We propose that accidental incorporation of self nucleic acids by CRISPR can incur an autoimmune fitness cost, and this could explain the abundance of degraded CRISPR systems across prokaryotes.

Modulation of intein activity by its neighboring extein substrates.

Inteins comprise a large family of phylogenetically widespread self-splicing protein catalysts that colonize diverse host proteins. The evolutionary and functional relationship between the intein and the split-host protein, the exteins, is largely unknown. To probe an association, we developed an in vivo and in vitro intein assay based on FRET. The FRET assay reports cleavage of the intein from its N-terminal extein. Applying this assay to randomized extein libraries, we show that the nature of the extein substrate bordering the intein can profoundly influence intein activity. Residues proximal to the intein-splicing junction in both N- and C-terminal exteins can accelerate the N-terminal cleavage rate by >4-fold or attenuate cleavage by 1,000-fold, both resulting in compromised self-splicing efficiency. The existence and the magnitude of extein effects require consideration for maximizing the utility of inteins in biotechnological applications, and they predict biases in intein integration sites in nature.

Bacterial gasdermins reveal an ancient mechanism of cell death.

Gasdermin proteins form large membrane pores in human cells that release immune cytokines and induce lytic cell death. Gasdermin pore formation is triggered by caspase-mediated cleavage during inflammasome signaling and is critical for defense against pathogens and cancer. We discovered gasdermin homologs encoded in bacteria that defended against phages and executed cell death. Structures of bacterial gasdermins revealed a conserved pore-forming domain that was stabilized in the inactive state with a buried lipid modification. Bacterial gasdermins were activated by dedicated caspase-like proteases that catalyzed site-specific cleavage and the removal of an inhibitory C-terminal peptide. Release of autoinhibition induced the assembly of large and heterogeneous pores that disrupted membrane integrity. Thus, pyroptosis is an ancient form of regulated cell death shared between bacteria and animals.

Multiple phage resistance systems inhibit infection via SIR2-dependent NAD+ depletion.

Defence-associated sirtuins (DSRs) comprise a family of proteins that defend bacteria from phage infection via an unknown mechanism. These proteins are common in bacteria and harbour an N-terminal sirtuin (SIR2) domain. In this study we report that DSR proteins degrade nicotinamide adenine dinucleotide (NAD+) during infection, depleting the cell of this essential molecule and aborting phage propagation. Our data show that one of these proteins, DSR2, directly identifies phage tail tube proteins and then becomes an active NADase in Bacillus subtilis. Using a phage mating methodology that promotes genetic exchange between pairs of DSR2-sensitive and DSR2-resistant phages, we further show that some phages express anti-DSR2 proteins that bind and repress DSR2. Finally, we demonstrate that the SIR2 domain serves as an effector NADase in a diverse set of phage defence systems outside the DSR family. Our results establish the general role of SIR2 domains in bacterial immunity against phages.

Bacteria deplete deoxynucleotides to defend against bacteriophage infection.

DNA viruses and retroviruses consume large quantities of deoxynucleotides (dNTPs) when replicating. The human antiviral factor SAMHD1 takes advantage of this vulnerability in the viral lifecycle, and inhibits viral replication by degrading dNTPs into their constituent deoxynucleosides and inorganic phosphate. Here, we report that bacteria use a similar strategy to defend against bacteriophage infection. We identify a family of defensive bacterial deoxycytidine triphosphate (dCTP) deaminase proteins that convert dCTP into deoxyuracil nucleotides in response to phage infection. We also identify a family of phage resistance genes that encode deoxyguanosine triphosphatase (dGTPase) enzymes, which degrade dGTP into phosphate-free deoxyguanosine and are distant homologues of human SAMHD1. Our results suggest that bacterial defensive proteins deplete specific deoxynucleotides (either dCTP or dGTP) from the nucleotide pool during phage infection, thus starving the phage of an essential DNA building block and halting its replication. Our study shows that manipulation of the dNTP pool is a potent antiviral strategy shared by both prokaryotes and eukaryotes.

An expanded arsenal of immune systems that protect bacteria from phages.

Bacterial anti-phage systems are frequently clustered in microbial genomes, forming defense islands. This property enabled the recent discovery of multiple defense systems based on their genomic co-localization with known systems, but the full arsenal of anti-phage mechanisms remains unknown. We report the discovery of 21 defense systems that protect bacteria from phages, based on computational genomic analyses and phage-infection experiments. We identified multiple systems with domains involved in eukaryotic antiviral immunity, including those homologous to the ubiquitin-like ISG15 protein, dynamin-like domains, and SEFIR domains, and show their participation in bacterial defenses. Additional systems include domains predicted to manipulate DNA and RNA molecules, alongside toxin-antitoxin systems shown here to function in anti-phage defense. These systems are widely distributed in microbial genomes, and in some bacteria, they form a considerable fraction of the immune arsenal. Our data substantially expand the inventory of defense systems utilized by bacteria to counteract phage infection.

Diversity and classification of cyclic-oligonucleotide-based anti-phage signalling systems.

Cyclic-oligonucleotide-based anti-phage signalling systems (CBASS) are a family of defence systems against bacteriophages (hereafter phages) that share ancestry with the cGAS-STING innate immune pathway in animals. CBASS systems are composed of an oligonucleotide cyclase, which generates signalling cyclic oligonucleotides in response to phage infection, and an effector that is activated by the cyclic oligonucleotides and promotes cell death. Cell death occurs before phage replication is completed, therefore preventing the spread of phages to nearby cells. Here, we analysed 38,000 bacterial and archaeal genomes and identified more than 5,000 CBASS systems, which have diverse architectures with multiple signalling molecules, effectors and ancillary genes. We propose a classification system for CBASS that groups systems according to their operon organization, signalling molecules and effector function. Four major CBASS types were identified, sharing at least six effector subtypes that promote cell death by membrane impairment, DNA degradation or other means. We observed evidence of extensive gain and loss of CBASS systems, as well as shuffling of effector genes between systems. We expect that our classification and nomenclature scheme will guide future research in the developing CBASS field.