A bacterium's sense of self explained

Foreigner or native-born? Your immune system discriminates between them, as do those of bacteria. Yes indeed, bacteria do have immune systems – pretty complex ones at that. And like any useful immune system, the bacterial ones must have a good technique for distinguishing “foreign” from “self.”

You may even have heard of the bacterial immune system: It’s called CRISPR, and it’s used in biology research around the world for DNA engineering and genome editing. CRISPR normally inserts short DNA sequences taken from phages – viruses that invade bacteria – into special slots called spacers within its genome. These bits of DNA form an “immune memory” – the record of past infection that helps fight the next one. The phage sequences are used as a template to create “antisense” RNA-protein complexes that can identify and take out further phages that try to sneak into the cell. This kind of immunity is adaptive and, until recently, scientists did not think that bacteria had something so sophisticated as an adaptive immune system.

bacteria cell and phages. Wikimedia commons

Within less than a decade of its discovery, researchers had revealed how CRISPR works and even found how to use it for other purposes, but there were still some fairly big open questions -- especially how it discriminates between foreign and self. How can it look at two sequences of DNA and know that one belongs to a phage, the other to its own genome? For a bacterium, this is quite critical: Mistakenly inserting a bit of self-DNA could cause a fatal “autoimmune” attack. But if the cell misidentifies the phage DNA as self, the results could be no less deadly. And the ID system must be nimble, as well, since most environments are home to many more phages than bacteria.

Prof. Rotem Sorek and his group, working with researchers at Tel Aviv University, “infected” bacterial cells with round bits of DNA called plasmids and then recorded some 38 million separate immunization events to see how and where the selection occurs. (That’s no mistake: They really have data from 38 million events.)

The CRISPR foreign-self discrimination mechanisms they discovered astonish and delight because they are clever and efficient, and they use the bacteria’s DNA copying machinery to do the job. Extra DNA stuck in the genome – say from a virus that wants to get replicated – will gum up the machinery as the DNA double strand is being unwound in preparation for copying. A stall in the process brings in the repair enzyme; this enzyme, along with several CRISPR-associated proteins, checks out the sequence.

In the end, it all comes down to differences in replication: Viral DNA, which pretty much exists to reproduce, will have a lot of genetic bits that replicate at high rates. What it won’t have much of is another sequence that is found all over that bacterial genome, which tells the copying machinery to stop. So if the repair machinery finds lots of one and doesn’t run into the other, it can “assume” the DNA comes from a phage.

Why should we care about a bacterial immune mechanism? In truth, many think we need to start taking more of an “ecosystem” approach to the whole subject of bacteria and viruses – our world is teeming with them. So, for example, if we want to use phages, as some have suggested, as antibiotics, we might have to understand how the bacteria could develop resistance. On the other hand, some of the first to investigate the commercial use of CRISPR have been the yogurt producers, who can lose their bacterial cultures to phages. Phages can also live symbiotically inside bacteria, and in the case of certain hosts, cause harmless bacteria to become pathogenic. And, of course, phages are part of your personal ecosystem – present in and alongside the microbiota that live in your gut and on your skin.

Sorek is now looking for other bacterial immune systems – CRISPR is found in only around half of all bacteria, so there are sure to be others. And that, he says, will open up a whole new set of questions about discrimination on the microscopic scale.