A Question of Function

I like blog carnivals. They expose me to stuff I might not otherwise see, and they expose people who might not see it to my stuff. Win-win. (Apart, if I’m permitted a niggle, from what seems to be a complete absence of carnivals devoted to agriculture. Is there any demand?) Anyway, the latest Mendel’s Garden had one post I found particularly interesting.

Keith Robison at Omics! Omics! was musing about restriction endonucleases. These are the molecular scissors that allow biologists to cut DNA accurately and repeatably, exactly where they want to. It’s fair to say that no modern molecular biology would be possible without them. By and large, restriction enzymes recognize a specific short stretch of DNA sequence and then break the DNA there. The big question, to me, is what are they for.

As I commented on Keith’s blog

I know they are useful tools and all, and nobody questions, for example, why the Swiss Army needs a spiffy knife. But I've seen very little speculation (apart from my own, ages ago) on the biological function of restriction enzymes. What do you think they are for (apart from being for molecular biologists)?

Keith replied:

I’m no expert, but the two general classes of evolutionary explanation are
1) Defend yourself against foreign genomes (e.g. viruses).
2) Restriction/methylation systems as selfish genetic systems which once acquired, can’t be evicted.
The wide variety of different RE architectures hints that there must be some interesting selection for them -- they keep being invented.

And he kindly linked to a review, which I promise to read when time permits. Meanwhile, I must say that explanation No. 1 cuts no ice with me, and never has. The idea is that some virus inserts its dastardly DNA into a bacterial cell. The cell’s restriction enzymes recognize their target sequences (if present) and promptly chop the invading DNA into useless little bits. Indeed, you could say that the bacterium is effectively eating the virus, reusing the precious DNA components. But hang on. What protects a piece of DNA from being cut up? Specific changes to the target sequence. And what if a virus acquires those modifications, just by chance? It survives, and proliferates. Of course I haven’t run the numbers, but I just do not believe that the bacterial restriction enzymes evolve quickly enough to keep up with the viruses.

That leaves the second option, which as it happens is the one I thought most likely way back in 1982, when I first scribbled about this stuff. In this scenario, the restriction enzyme is anti-sex. Yes, bacteria do exchange their DNA. And if DNA from another individual does not have the same restriction enzyme, then the host is once again going to break it up and recycle its parts.

But why? Because it can. Restriction enzymes might just be the ultimate selfish DNA, with just one function: to destroy anything that isn’t a copy of itself.

Like I said, I’ll read that review soon. But it is actually rather hard to get up to speed on the topic. That’s because when an evolutionary biologist wants to know what something is for, she tends to ask what is its function. Why was it selected? What advantage did it confer? When a molecular biologist uses the word “function” there’s usually a different underlying question: how does it work? And many more people seeming to be asking the how question than the why question. They’re both perfectly good questions, of course, but Google can't tell the difference between them. And therefore neither can I.

Gosh but it would be good to be able to pursue these thoughts properly.