Structural modeling reveals phage proteins that manipulate bacterial immune signaling.
Saved in:
| Title: | Structural modeling reveals phage proteins that manipulate bacterial immune signaling. |
|---|---|
| Authors: | Tal, Nitzan (AUTHOR), Hadary, Romi (AUTHOR), Chang, Renee B. (AUTHOR), Osterman, Ilya (AUTHOR), Jacobson, Roy (AUTHOR), Yirmiya, Erez (AUTHOR), Bechon, Nathalie (AUTHOR), Hochhauser, Dina (AUTHOR), López Rivera, Miguel (AUTHOR), Madhala, Barak (AUTHOR), Garb, Jeremy (AUTHOR), Goldsmith, Moshe (AUTHOR), Wein, Tanita (AUTHOR), Kranzusch, Philip J. (AUTHOR), Amitai, Gil (AUTHOR), Sorek, Rotem (AUTHOR) |
| Source: | Science. 3/5/2026, Vol. 391 Issue 6789, p1-16. 16p. |
| Subjects: | Structural models, Viral proteins, Immunity, Cyclic nucleotides |
| Abstract: | Immune systems in animals, plants, and bacteria often rely on intracellular nucleotide signaling, which viruses can block by sequestering or degrading these signals. We identified structural and biophysical traits shared by diverse viral antidefense proteins and used these traits to develop a computational pipeline that predicts phage proteins whose role is to manipulate bacterial immune signaling. Experimental validation revealed three previously uncharacterized protein families—Sequestin, Lockin, and Acb5—that inhibit the Thoeris system and the cyclic oligonucleotide–based antiphage signaling system (CBASS). Sequestin and Lockin act as nucleotide "sponges," binding 1″–3′ glycocyclic adenosine diphosphate–ribose (3′cADPR) and histidine conjugated to ADPR (His- ADPR), whereas Acb5 cleaves cyclic guanosine monophosphate–adenosine monosphosphate (3′3′-cGAMP) and related molecules. Structural and mutational analyses explain their binding and catalytic mechanisms. Thousands of homologs occur in phage genomes, highlighting the abundance and diversity of viral strategies to subvert nucleotide-based immunity. Editor's summary: When a bacterium is infected by a phage virus, it uses small signaling molecules to activate its defenses. Phages, in turn, have evolved ways to evade these defenses by capturing and "confiscating" the alert molecules, but how they do so has remained largely unknown. Tal et al. used artificial intelligence–based approaches to scan vast collections of phage proteins and identify factors that sequester or degrade immune signaling molecules (see the Perspective by Fedorova and Bondy-Denomy). The proteins they uncovered are found in thousands of phages, including the classic model phage T4 that has been studied for decades, and explain how phages overcome bacterial immune systems. Looking ahead, the approach described here could be applied to discover analogous proteins in viruses that infect animals and plants. —Di Jiang INTRODUCTION: Many innate immune pathways in bacteria, plants, and animals use nucleotide derivatives as intracellular immune signaling molecules. Phages have evolved counterdefense "sponge" proteins that inhibit bacterial immune signaling by binding and sequestering the immune signals as well as enzymes that can cleave and inactivate the signaling molecules. A few antidefense sponge and enzyme protein families were recently discovered serendipitously, suggesting that they are abundant among phages. Our goal was to develop an unbiased approach to computationally predict and experimentally verify previously unknown families of antidefense enzymes and sponge proteins within vast databases of viral protein sequences. RATIONALE: We reasoned that, despite their divergence in sequence and structure, viral proteins that target nucleotide signals might share common structural and biophysical traits arising from their shared function. By comparing the structures of known viral sponge proteins, we identified unifying features: they are small (typically ≤100 amino acids); form homo-oligomeric assemblies; and have deep, positively charged pockets that bind the negatively charged nucleotide signals. We also found that sponges from different families are frequently encoded fused as a single polypeptide. We hypothesized that these structural fingerprints could be used to discover viral sponges through structural modeling of large sets of viral proteins. RESULTS: We built a structure-guided computational pipeline that searches viral protein databases either for proteins fused with known sponges or for small, unannotated proteins predicted by AlphaFold-Multimer to form homo-oligomers with internal positively charged pockets. Screening a clustered database of 32 million proteins from 2 million phage genomes resulted in >120 candidate proteins that were experimentally tested for their ability to inhibit bacterial immune signaling. This resulted in the discovery of three previously unidentified families of phage-encoded antidefense proteins. Two families, Sequestin and Lockin, are sponges that sequester the Thoeris immune signaling molecules 1′′–3′ glycocyclic adenosine diphosphate–ribose (3′cADPR) and histidine conjugated to ADPR (His-ADPR). A third family, Acb5, was found to inhibit cyclic oligonucleotide–based antiphage signaling system (CBASS) immunity by enzymatically degrading its signaling molecule, cyclic guanosine monophosphate–adenosine monosphosphate (3′3′-cGAMP). Structural modeling, crystallographic analyses, and biochemical assays confirmed the computational predictions and explained how these proteins neutralize the bacterial immune signals. Homologs of these antidefense protein families were found encoded in thousands of viral genomes, including in well-studied phages such as T2, T4, and T6, which suggests that manipulation of host immune signaling is a common and evolutionarily conserved viral strategy. CONCLUSION: Our work introduces a structure-based discovery framework for detecting viral proteins that manipulate host immune signaling, independent of sequence similarity or prior annotation. The identification of previously undescribed families of viral sponges and nucleotide-cleaving enzymes in thousands of phage genomes demonstrates that these counterdefense strategies are pervasive in the phage world. This approach could be used in the future to discover analogous inhibitors in viruses infecting any organism of choice, including eukaryotes. Structure-guided discovery of phage inhibitors of bacterial immune signaling.: Small phage proteins of unknown function were modeled using AlphaFold and screened for oligomeric assemblies containing positively charged pockets. Experimental validation revealed three previously uncharacterized protein families: Sequestin and Lockin, sponges that inhibit the Thoeris defense system, and Acb5, an enzyme that cleaves CBASS-produced signals. This study establishes a computational approach for unbiased discovery of immune modulators across phage genomes. cUA, cyclic uridine monophosphate–adenosine monosphosphate. [ABSTRACT FROM AUTHOR] |
| Copyright of Science is the property of American Association for the Advancement of Science and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.) | |
| Database: | Psychology and Behavioral Sciences Collection |
|
Full text is not displayed to guests.
Login for full access.
|
|
| Abstract: | Immune systems in animals, plants, and bacteria often rely on intracellular nucleotide signaling, which viruses can block by sequestering or degrading these signals. We identified structural and biophysical traits shared by diverse viral antidefense proteins and used these traits to develop a computational pipeline that predicts phage proteins whose role is to manipulate bacterial immune signaling. Experimental validation revealed three previously uncharacterized protein families—Sequestin, Lockin, and Acb5—that inhibit the Thoeris system and the cyclic oligonucleotide–based antiphage signaling system (CBASS). Sequestin and Lockin act as nucleotide "sponges," binding 1″–3′ glycocyclic adenosine diphosphate–ribose (3′cADPR) and histidine conjugated to ADPR (His- ADPR), whereas Acb5 cleaves cyclic guanosine monophosphate–adenosine monosphosphate (3′3′-cGAMP) and related molecules. Structural and mutational analyses explain their binding and catalytic mechanisms. Thousands of homologs occur in phage genomes, highlighting the abundance and diversity of viral strategies to subvert nucleotide-based immunity. Editor's summary: When a bacterium is infected by a phage virus, it uses small signaling molecules to activate its defenses. Phages, in turn, have evolved ways to evade these defenses by capturing and "confiscating" the alert molecules, but how they do so has remained largely unknown. Tal et al. used artificial intelligence–based approaches to scan vast collections of phage proteins and identify factors that sequester or degrade immune signaling molecules (see the Perspective by Fedorova and Bondy-Denomy). The proteins they uncovered are found in thousands of phages, including the classic model phage T4 that has been studied for decades, and explain how phages overcome bacterial immune systems. Looking ahead, the approach described here could be applied to discover analogous proteins in viruses that infect animals and plants. —Di Jiang INTRODUCTION: Many innate immune pathways in bacteria, plants, and animals use nucleotide derivatives as intracellular immune signaling molecules. Phages have evolved counterdefense "sponge" proteins that inhibit bacterial immune signaling by binding and sequestering the immune signals as well as enzymes that can cleave and inactivate the signaling molecules. A few antidefense sponge and enzyme protein families were recently discovered serendipitously, suggesting that they are abundant among phages. Our goal was to develop an unbiased approach to computationally predict and experimentally verify previously unknown families of antidefense enzymes and sponge proteins within vast databases of viral protein sequences. RATIONALE: We reasoned that, despite their divergence in sequence and structure, viral proteins that target nucleotide signals might share common structural and biophysical traits arising from their shared function. By comparing the structures of known viral sponge proteins, we identified unifying features: they are small (typically ≤100 amino acids); form homo-oligomeric assemblies; and have deep, positively charged pockets that bind the negatively charged nucleotide signals. We also found that sponges from different families are frequently encoded fused as a single polypeptide. We hypothesized that these structural fingerprints could be used to discover viral sponges through structural modeling of large sets of viral proteins. RESULTS: We built a structure-guided computational pipeline that searches viral protein databases either for proteins fused with known sponges or for small, unannotated proteins predicted by AlphaFold-Multimer to form homo-oligomers with internal positively charged pockets. Screening a clustered database of 32 million proteins from 2 million phage genomes resulted in >120 candidate proteins that were experimentally tested for their ability to inhibit bacterial immune signaling. This resulted in the discovery of three previously unidentified families of phage-encoded antidefense proteins. Two families, Sequestin and Lockin, are sponges that sequester the Thoeris immune signaling molecules 1′′–3′ glycocyclic adenosine diphosphate–ribose (3′cADPR) and histidine conjugated to ADPR (His-ADPR). A third family, Acb5, was found to inhibit cyclic oligonucleotide–based antiphage signaling system (CBASS) immunity by enzymatically degrading its signaling molecule, cyclic guanosine monophosphate–adenosine monosphosphate (3′3′-cGAMP). Structural modeling, crystallographic analyses, and biochemical assays confirmed the computational predictions and explained how these proteins neutralize the bacterial immune signals. Homologs of these antidefense protein families were found encoded in thousands of viral genomes, including in well-studied phages such as T2, T4, and T6, which suggests that manipulation of host immune signaling is a common and evolutionarily conserved viral strategy. CONCLUSION: Our work introduces a structure-based discovery framework for detecting viral proteins that manipulate host immune signaling, independent of sequence similarity or prior annotation. The identification of previously undescribed families of viral sponges and nucleotide-cleaving enzymes in thousands of phage genomes demonstrates that these counterdefense strategies are pervasive in the phage world. This approach could be used in the future to discover analogous inhibitors in viruses infecting any organism of choice, including eukaryotes. Structure-guided discovery of phage inhibitors of bacterial immune signaling.: Small phage proteins of unknown function were modeled using AlphaFold and screened for oligomeric assemblies containing positively charged pockets. Experimental validation revealed three previously uncharacterized protein families: Sequestin and Lockin, sponges that inhibit the Thoeris defense system, and Acb5, an enzyme that cleaves CBASS-produced signals. This study establishes a computational approach for unbiased discovery of immune modulators across phage genomes. cUA, cyclic uridine monophosphate–adenosine monosphosphate. [ABSTRACT FROM AUTHOR] |
|---|---|
| ISSN: | 00368075 |
| DOI: | 10.1126/science.aea1761 |
