If you are a bacterium, life can be pretty dangerous. Wherever you live, you are surrounded by millions of viruses that can inject you with their DNA and turn you into a helpless factory of these (bacteriophage T4 by Adenosine, Wikipedia):
(They may kill you, but at least they look badass…)
In response to the constant threat of deadly viruses, bacteria and archaea have evolved an ingenious defence mechanism called the CRISPR system. The CRISPR region in the genome consists of repetitions of a short sequence (the actual CRISPRs), alternating with spacers containing foreign (often viral) DNA the bacterium snatched from invaders. Spacers can be pulled out later and used to recognise and destroy the same invader.
If you are a CRISPR-enabled cell and survive a virus attack long enough to add the attacker to your “library”, you and your offspring are protected against the same virus forever. (OK, in reality, if a virus doesn’t show up for many generations, the microbes can afford to lose their “memory” of it to mutations. But provided that the virus is around and exerting a selection pressure, the immunity remains.)
The CRISPR system is fascinating for more than one reason. First, it’s essentially an adaptive immune system in some of the simplest organisms alive today. Adaptive immune systems are uncommon even in multicellular creatures. They are known from two groups of vertebrates (jawless and jawed vertebrates probably evolved them independently) and possibly brown algae (Zambounis et al., 2012).*
Second, it’s eerily “Lamarckian”. Microbes with CRISPR can, in a way, “direct” their own evolution. They are acquiring new adaptations over their lifetimes, and passing these on to their descendants. And they aren’t doing it in the way most inheritance of acquired traits works – instead of tagging their DNA with signals that remain highly flexible in the long term, they are actually permanently incorporating the new information into their genomes.
Parasites are notorious for evolving a way around anything a host can throw at them, though, and bacteriophages are no exception. A new study not-quite-published in Nature (Bondy-Denomy et al., 2012) reports viral genes that enable the viruses to break through CRISPR-based defences. They experimented on cultures of a strain of the bacterium Pseudomonas aeruginosa. Their bacteria went into the experiment infected with a variety of viruses, but the viruses were kept in their dormant form that doesn’t hurt the host cell. When these bacteria were exposed to active viruses, most of them could defend themselves just fine, while genetically engineered bacteria who lack a CRISPR system die like flies from the exact same pathogens.
However, a few of the bacteria were unable to resist the assault despite a fully intact immune system. Since the only difference between the different bacterial samples was the kind of dormant virus they hosted, the reason had to be something related to the viruses. A look at the viral genomes turned up several genes that almost literally held smoking guns. When added to CRISPR-sensitive viruses, they enabled them to kill bacteria they couldn’t otherwise harm. When deleted from their source viruses, they prevented the viruses from killing CRISPR-enabled but not CRISPR-disabled bacteria. A series of such experiments demonstrated that these “anti-CRISPR” genes helped viruses evade the immune systems of their hosts. Interestingly, they didn’t work against the closely related CRISPR system of E. coli – the cheat codes, so to speak, were highly specific to the game.
A really interesting thing about the CRISPR resistance genes is their possible origin. It seems Pseudomonas may have wrought its own doom in this case! When the researchers searched for sequences similar to the resistance genes, some of the sequences they caught were actually from other strains of Pseudomonas itself. A possible explanation is this: since CRISPR-based immunity can work on any foreign DNA, the original benefit of some anti-CRISPR genes may have been to prevent the destruction of bacterial genes after being passed to other bacteria. Then the viruses somehow got their hands on the sequences, hacking poor microbes with their own code.
Evolutionary arms races are such weird and crazy and fascinating things!
*I can’t really figure out whether some RNA interference based viral defence in plants counts. I’d have to go fact-hunting to find out if any of the interfering RNAs originate in a similar way to CRISPR spacers. The whole adaptive immunity angle is a digression, though, so I’m lazy enough to leave that up in the air.
Bondy-Denomy J et al. (2012) Bacteriophage genes that inactivate the CRISPR/Cas bacterial immune system. Nature advance online publication, available 16/12/2012, doi: 10.1038/nature11723
Zambounis A et al. (2012) Highly dynamic exon shuffling in candidate pathogen receptors… what if brown algae were capable of adaptive immunity? Molecular Biology and Evolution 29:1263-1276