I have a good friend who has just emerged triumphant from a run-in with bowel cancer – she’s in complete remission! Almost as wonderful is the fact that colliding with cancer has converted her from a genuine non-scientist to one who devours biology like fish and chip suppers. Spotting a recent volley of media items about four-stranded ‘quadruple helix’ DNA in human cells, she was on Twitter in a flash: “Does this mean that people with cancer have lots more quadruplex DNA than normal?” As she knows I can’t stand the Tweet cult she was probably amazed to get a reply: short answer: “No.”
But as ever in science, there’s a long(er) response. So, if you’re interested in the gyrations and gymnastics of which your genetic code is capable, read on …
As you know, DNA comes as a double helix – a 2-chain spiral of small units (called nucleotides) that stick together (the units contain bases, so they’re ‘base-paired’). The oft-reproduced double helix image is beautiful because it’s a repetitive structure and you can easily see how it can be ‘unzipped’ so that each half can be used as a template to make a copy and regions can be ‘read’ to make RNA and proteins – though it was really designed to enable biologists to make endless unzipping jokes about genes and jeans.
The two DNA molecules of the helix stick together because of a balance between three forces: (1) weak electronic attraction between some atoms in the bases (called hydrogen bonds), (2) a sort of glueyness between the bases because their chemical structure means they don’t like water much and they’d rather snuggle up together, and (3) a repulsion between the chains because of the repeated phosphate groups all the way along the backbone (these carry an electric charge and likes repel, as we know).
But with all these attractions and repulsions you might think there would be lots of ways nucleic acids could get tangled up with each other – and there is. So the common form of double helix (the beautiful shape) is B-DNA but there’s also A-DNA, C-DNA and Z-DNA. If you just change the conditions a bit (pinch of salt or whatever) you can tweak the interactions so bits of the bases that don’t interact in B-DNA will do so to give a slightly different shape (usually a bit distorted – ugly). As you can see from the structure, Z-DNA is more Homer Simpson than Watson and Crick.
B-DNA and Z-DNA
We can’t reproduce the environment of DNA in the nucleus so we don’t really have a clue but the betting is that short bits of DNA jink in and out of these odd structural formations – just as part of the continuous flexing of the molecules. There’s also a couple of other things that can happen that have been known for a long time – again just dependent on the precise conditions in which a piece of DNA finds itself.
The first is a variant on the hydrogen bonds that form between bases. One way to think of this is to imagine two circles of five people, each ring holding hands and facing outwards. Each person is an atom in the bases of DNA. Let’s think of base pairing as the two nearest in the circles getting close enough to kiss. That’s one hydrogen bond. But, of course, the two other pairs on either side will now be quite close: if one of them also manages to kiss (tongues may be used) now we have two hydrogen bonds – which is what holds the bases A and T together. But suppose that the pair on the other side (who must also be quite close with all this adjacent necking going on) decide they really fancy joining in and are so excited that they twist the circle out of shape to do so. That this can happen has been known for years (it’s called Hoogsteen base pairing after the voyeur what spotted it) and when it does it can distort the helix enough for a third DNA strand to wrap round the original two – so you get triple-stranded DNA.
A sexual need
Similarly, if you tweak the conditions you can get four strands of DNA to come together and indeed we’ve known for yonks that happens naturally during recombination (that’s when genetic material gets swapped between Mum and Dad chromosomes – the reason for sex). When that happens you can think of four DNA strands forming a cross, each quadrant contains DNA from one strand of a chromosome, base-paired to that in the next quadrant – which is how bits get swapped around.
So there’s nothing new about odd DNA shapes but what has made the news is that for the first time, rather than looking at what can be made to happen in a test tube, Shankar Balasubramanian and his pals have looked in whole cells. To do this they made an antibody that sticks only to ‘quadruple helix’ DNA structures – G-quadruplexes. The upshot is that they detected quadruplexes scattered throughout chromosomes and they see more in cells that are rapidly dividing than in ones that are just sitting there (they looked in some cancer cells in culture that do divide quite rapidly – but bear in mind that in tumours cells aren’t diving all that fast). So the inference is that they might form as part of DNA replication and, if you can target them by their antibody, maybe you could do something similar with a drug that would stop cells dividing. And if you could target that to cancer cells you could stop them in their tracks.
And the catch …
Simple. But there are some problems. It’s possible the antibody helps the quadruplexes to form – so it could even be a cunning artifact. But if we assume it isn’t – then we come face to face with a really big problem. There are zillions of ways you can kill cancer cells. The difficulty is that there isn’t one that selects cancer cells from normals. It may be possible (though it’s not evident how) to target quadruplexes and block cell division – but there are lots of cells that we need to divide rapidly just to keep us going – and, if quadruplexes are real, presumably they have ’em. So non-specific killing is probably not a good idea. Twas ever thus.
Biffi, G., Tannahill, D., McCafferty, J. and Balasubramanian, S. (2013). Quantitative visualization of DNA G-quadruplex structures in human cells. Nature Chemistry published online: 20 January 2013 | doi: 10.1038/nchem.1548