A follow-up on the "crazy stack" story.
(work in progress)
A database about base pairs
FRABASE classifies this pair as a Sänger XI, which corresponds to BPS pattern GA_24.
If I'm not mistaken, this is the classical sheared G.A pair. There are 2 other GA pairs in the loop. The one at the other end of the loop (A78-G98) is structurally identical to this one. But the one in the middle of the loop is very different.
FRABASE classifies this pair as a Sänger XXIV (nicknamed "AU Reversed Hoogsteen"), which corresponds to BPS pattern AU_14.
This is a non-canonical AU pair. Compare with BPS pattern AU_2, which is the canonical Watson-Crick one.
And again, the other U77-A99 pair is identical in shape to this one.
FRABASE doesn't associate this pair with a Sänger class, and the best match I could find is BPS pattern GU_23.
This is not the GU Wobble, which is BPS pattern GU_1
FRABASE doesn't associate this pair with a Sänger class, and the best match I could find is BPS pattern GA_94, which is different than the two other sheared G.A ones present in this loop.
FRABASE classifies this pair as a Sänger VII, which corresponds to BPS pattern GG_21, but I think this might be a mistake. I measured the distances between hydrogen donors and acceptors (visible in the picture), and the only valid possibilities just don't seem to match with the proposed pattern...
Isosteres are base pairs with similar geometries. In other words, they can replace one another without structural distortion.
Well, I've been wondering about replacing some of the pairs of the "crazy stack" with isosteres, and compare results. Let's review.
- I'd rather leave the canonical pairs for another occasion.
- G72-A104, as already mentioned, is a BPS GA_24 pattern, which belongs to the GA_24 isostere class. Now the question is: can a CU pair (CU_33 is presented as "compatible") really replace a GA one? I kinda get the point that the positions and distances of the C1'-C1' doublet is in a very acceptable range (which means that the backbone shouldn't be influenced much), but I have some strong doubts about the stacking interactions...
The other possibly interesting experiment would be to replace this GA by a AA one, hoping to see it form the BPS AA_14 pattern, which is a priori compatible. This time the geometry should be fully compatible, even for stacking, since we're exchanging a purine for another. Though, a surprise may come from the hydrogen-bonding topology. In the case of this AA pair, one of them links to the O2' end on the ribose, which is precisely the point where the SHAPE reagent probes the nucleotide. What's gonna happen? Will the probe break this specific bond, actually destabilizing the pair and making the base look reactive? Or will it appear that this A base is even better protected from the SHAPE probe?
- Similarly, the A73-U103 reversed Hoogsteen is the template of the AU_14 isostere class. Among the possible replacements listed, the AC one (AC_48) could be interesting. As in compatible, while probably weaker, which may prove useful in some contexts.
- BPS doesn't offer alternatives for the central 3 pairs, so I guess that would be the end of possible experiments based on changing single pairs or nucleotides... This said, when you consider the pictures for U74-G102 and G76-G100 side by side, doesn't it feel like they share a lot of features?