You have full access to this article via Jozef Stefan Institute.
New components of bacterial defence systems are rapidly being discovered. The average bacterium encodes seven systems that target invaders such as bacteria-infecting viruses, which are known as bacteriophages or phages. However, the question remains whether these mechanisms work together to form an immune system that is more than the sum of its parts. Writing in Nature, Doherty et al. and Sullivan et al. report the discovery of an antiviral defence system that was named Panoptes, after the many-eyed giant of Greek mythology who was a watchman for the gods. This defence specifically targets viruses that can evade another common bacterial defence system, revealing a sophisticated and layered antiviral protection network for bacteria.
Bacterial defence mechanisms are incredibly diverse, with hundreds of them already identified. One common mechanism is the generation of molecules called cyclic-nucleotide second messengers in response to viral infection. An example of this is CBASS, which evolved into the cGAS-STING defence pathway found in cells with nuclei (eukaryotic cells). When viral infection is sensed, an enzyme called CD-NTase generates a type of cyclic nucleotide called a cyclic dinucleotide. These molecules bind to and activate a wide range of proteins that provide antiviral defences by targeting invading viruses or by harming or killing the bacterial cell to prevent phage replication. For example, activation of a phospholipase enzyme can kill cells by causing membrane disruption.
In response to this, phages have evolved anti-defence proteins that interfere with the signalling mediated by cyclic dinucleotides. The most notable of these is the 'sponge' proteins that bind to the cyclic dinucleotides. CBASS systems can generate a wide variety of cyclic dinucleotides, and viral sponges have, in response, evolved multiple binding pockets for these molecules and have increased their binding capacity to 'jam' the signals.
At the heart of the newly identified Panoptes system is the ability of the bacteria to generate decoy cyclic nucleotides, which when they are 'mopped up' by invading viruses, provide a mechanism to alert bacteria to viral infection (Fig. 1). An enzyme named mCpol, which is related to the Cas10 subunit of a bacterial defence system called type III CRISPR, generates cyclic dinucleotides of a type called cyclic-oligoadenylate second messengers. Both Doherty et al. and Sullivan et al. show that mCpol is constantly active, producing an unusual type of cyclic oligoadenylate -- cyclic di-adenosine monophosphate (c-di-AMP).
This molecule binds to a membrane-bound protein related to the Cap15 protein of CBASS systems. Cap15 proteins can disrupt cell membranes to provide a defence response by killing the infected bacterial cell. The CBASS Cap15 is activated by binding to a cyclic dinucleotide, whereas the Panoptes version of Cap15 is inhibited by cyclic dinucleotide binding. Therefore, phages entering a cell that is armed with a Panoptes defence system will trigger a response if they attempt to use a sponge protein to disrupt cyclic-dinucleotide signalling. This is because c-di-AMP depletion will derepress the cell's version of Cap15.
Although the Panoptes system provides antiviral defence on its own, its real power comes when it works in tandem with CBASS. In such a scenario, viruses without sponges activate CBASS, whereas those armed with sponges activate Panoptes. Although this seems like an ideal arrangement for a bacterium, Panoptes is a rare system and most CBASSs are encoded in genomes that lack Panoptes.
A cyclic dinucleotide must satisfy several conditions to be used by a Panoptes-type system. It must be resistant to degradation by enzymes called phosphodiesterases, it must not interfere with any other cellular signalling pathways and yet be bound to and sequestered by viral sponges that could presumably evolve to ignore it. Notably, the c-di-AMP dinucleotide made by mCpol in these studies has a type of bond (a 2'-5' phosphodiester bond) that is rare in bacteria. Furthermore, any disruption of Panoptes signalling might result in toxicity and self-destruction. This combination of limitations might mean that Panoptes defence systems evolve only in situations in which viral sponges are frequently encountered.
Although the two studies focused on defence systems using a Cap15-like protein, it is known that other defence systems couple mCpol with a CRISPR-associated CARF protein. Sometimes, the two proteins are even fused to form a single multi-domain protein. The association of a Cas10-like enzyme with a CRISPR-like response protein prompts speculation that Panoptes is ancestral to the widespread type III CRISPR defence system found in bacteria.
Panoptes joins the growing list of antiphage defence systems that detect viral attempts to thwart bacterial defences. Another notable example is termed the PARIS defence system, which activates when phages inhibit bacterial enzymes that can target invading DNA. When coupled with the observation that several antiphage defence systems can function much better cooperatively than they can alone in bacteria, scientists can confidently conclude that bacterial defences are indeed more than just the sum of their parts.