Even an unevenly-rolled ribbon still compactifies more than a helix.
Showing exactly how the genetic information is recorded was central to our understanding of molecular biology.
One thing that was clear right away in 1953 was the significance of a strucure with two strands annealed by base-pairing: A on one strand always opposite T on the other, G opposite C. So each strand encodes the same information (one is the “reverse complement” of the other). In Watson & Crick’s structure, the base-pairing is mediated by hydrogen bonds, which are the strongest non-covalent bonds. Secure hydrogen bonding requires that the correct atoms “line up” opposite one another, so the double-stranded helix is highly stable only when base-pairing occurs, i.e. only when the correct complementary bases are opposite one another, when the same information is on both strands. However, since hydrogen bonds are still weaker than covalent bonds, the structure can be “unzipped” into single strands without compromising the integrity of the covalent backbone of each strand.
This was immediately suggestive of a replication mechanism whereby the two strands would be “unzipped”, and each single strand could serve as a template for the creation of a new complementary strand. Because Watson & Crick were too sexy for their shirts, they did not spell this out with great fanfare and excitement, but just inserted an understated single sentence:
http://www.nature.com/nature/dna50/watsoncrick.pdf
But the structure was a starting point for modern molecular biology, and many more aspects of structure & function have since emerged, for example:
Most important, the information in DNA is accessed dynamically. A required segment of DNA information is first copied to a much shorter intermediate, single-stranded messenger RNA, a similar nucleic acid polymer, again exploiting the unzipping and base-pairing feature. The location and timing of readout is controlled by complex regulatory networks, principally protein enzymes called transcription factors. Those that bind directly to DNA exploit the structural characteristics of the helix, notably that fact that the helical grooves in between the backbones expose the interior bases somewhat so that transcription factors can bind to specific sequences without needing to unzip the helix.
As discussed earlier in the thread, the basic overall structure of DNA is not sequence-dependent, because although the 4 bases are individually quite different in size and shape, when paired up the A-T pair and the G-C pair are similar in size and shape. This is not often emphasized, but it is interesting and important how nature has solved the problem of making its data storage medium the same overall size and shape whatever information is stored on it.
Base-pairing is also critical to fidelity, since some kinds of mutations can be repaired by checking against the opposite strand. A cool feature is that it turns out that DNA repair enzymes can “flip out” individual bases by rotating them 180-degrees around the axis of the backbone of their particular strand to the outside of the helix to inspect them.
That was a bit technical for my wee non-biochemist brain but I think I got the gist of it. Thanks!
There’s no way to roll a flat ribbon without spinning either the free end of the ribbon or the reel itself. Both of these become impractical if the ribbon is long enough (imagine reeling a mile-long ribbon).
There is a way to coil things where the odd coils subtract the twist while the even coils add, leading to a net zero, but it’s hard to imagine a molecule using this technique without helpers.
Plenty of processes introduce excess coiling or uncoiling of DNA, but this can neatly be resolved by cutting the backbone of one strand (by enzymes called topoisomerases). The cut strand spins freely to relieve any stress, and once they’re back at a neutral position the backbone can be reconnected.
There’s no reason why a flat ladder couldn’t have a similar process. Really, any linear polymer would need a way to deal with twisting and knotting.
Sure, but remember again that xNA was selected very early on–before there was a whole suite of biochemical helpers available. So while there are workarounds for all this stuff, those workarounds themselves have to be evolved-for, and that could give the advantage to something that already has a twist to it.
Or not. I’m really just arguing that the helix being a selected-for attribute is plausible, not that it had to be one.
I think the question of whether helix was selected over non-helix is probably not an important point for us to belabor.
There is surely little doubt that the structure and stability characteristics of the particular helix of the genetic material were selected and fine-tuned by evolution, since varying the molecular details of the nucleic acid polymer will influence aspects of the structure. Whether the pool of variation from which dsDNA was selected included non-helical structures or just a variety of helical structures isn’t really all that important.
The more interesting question is what factors were important in early natural selection for the genetic material. I’ve argued that a twofold compaction effect is not important today, but you have made the good point that it may have been more important in the early stages before the other more sophisticated compaction mechanisms evolved. As for other factors - today, the dsDNA helix draws a carefully optimized balance between stable preservation of information and accessibility for transcription and ease of replication. In the early days before most enzymes evolved, fine-tuning the temperature-dependent stability characteristics of the helix may have been critical, since readout and replication may have exploited diurnal or local temperature variations to denature and re-anneal the genetic material. I know the abiogenesis literature has a lot more on this, but I’m not up to date on it - I’ll try to see what I can find.
How about folding it back and forth? (Accordion fold.)
(I can see some self-interference problems, as bonds might try to form across the folds… But isn’t that also true for flat coils?)
Maybe–but as Riemann said, DNA is quite stiff. Of course we don’t know how stiff an alternative would be, but it seems likely that you don’t want the strand to be too floppy–it needs to feed properly into various other bits of machinery.
Perhaps you could have “hinges” that allowed it to fold in particular places easily. This seems to have its own set of problems, though.
I was thinking at first that this wouldn’t be very scalable, but then I realized that a long accordion-fold is itself a ribbon shape, and then you could accordion-fold that, and then recursively up to whatever level you need. So it would have that in common with the spiral, but it still doesn’t seem to have the same packing efficiency assuming that the folds can’t be creased flat.