I've been (mis)using this wiki as an account of my explorations in the RNA world. If nothing else, this specific trip inspired me this "philosophical" thought: sometimes, as you arrive at your destination, you realize that you can't find anything better in that place, than where you started from. That doesn't mean though, that the trip was worthless, if you learned something along the way...
After inputting the parameters in the search form, a miracle: there's a hit!
I would just have had to copy and paste that sequence in the lab designer, and call it a day. But I did something quite stupid...
I noticed the 'Show base pairs' button. That was already bad, but it got worse: I clicked it.
And I ended up with a lot more information than I had originally bargained for (ha, serves me right)
Well, once I had the data, I had to look at it, right? and so I did. Unfortunately, it was once again impossible to miss that beside the expected non-canonical interactions within the loop itself, there was a large number of "external" influences as well.
It was too late to back down, a 3D exploration was necessary.
Homology model for the Spinach chloroplast 50S subunit fitted to 9.4A cryo-EM map of the 70S chlororibosome
If you're like me, you've understood a whopping 5% of previous sentence: spinach. Congratulations!
Well, there's a first time for everything, vegetables should prove a refreshing change from viruses, bacteria, rice and human genome.
Ok, (spinach-flavored) popcorn time. Enjoy :)
The "loop"internal loop. Obviously, it looks much more like a double helix than say a balloon, but it's called a loop.
So here, a first remark about scientific terminology: if a base is paired canonically (or implicated in a GU wobble), then it is part of a stack, in all other cases, that base is part of a loop. This includes cases where for instance, an A is paired with a U, but not in the canonical cis-Watson-Crick/Watson-Crick conformation. Scientists will still call that a loop.
In the rest of this document, I will refer to this loop by the letter « X », and something like « A-x » means "the previously mentioned Adenine in the X loop".
The hairpinhairpin looks like a stack, capped by an apical loop. Though, there seem to be a few interesting features, in the form of apparent points of contact with the X loop...
By order of appearance
Wherever this large RNA structure came from, it was built in a specific order. The X loop was transcribed first, the Y hairpin somewhat later. Then X and Y got entangled in various interactions, and since both belong to the same sequence, these are called tertiary interactions.
At a later point in time, another sequence is transcribed, the Z stack forms, and in a final step, X and Z interact too. Since they belong to different sequences, those interactions are called quaternary ones.
Tertiary interaction I
Left picture: the Cytosine and the Guanine in the background, both part of the Y hairpin, are viewed from their aromatic ring plane. Both are paired, not with each other though. The remarkable feature is that the Adenine from the X loop (which I'll refer to as A-x from now on), is at about 45° inclination relative to the other bases.
Right picture: the purple marked atoms represent the possible endpoints for hydrogen bonds. So, it seems that A-x is "connected" to the two other bases.
Tertiary interaction II
This time, an Adenine of the Y hairpin loop is apparently inserted inside the stacking area of the X internal loop. This A-y seems to be paired with another A in the X loop, and to provide stacking support for bases above and below
Left picture: here we can see a repeated pattern: the Adenines of the X loop are positioned at a diagonal angle relative the the planes of the pairs in the Z stack.
Right picture: in purple, a large number of potential endpoints for hydrogen bonds.
A few ponderings
Not sure about you guys, but if any of you has given a little more than a quick glance to a scaffolding, you should have noticed these diagonal bars called cross braces. Do you know what they are for? Does this evoke anything in relation to some of the pictures above?
Let's imagine for a moment that the A-X / Y hairpin in Tertiary Interaction I just doesn't exist. Would A-y be able to find the keyhole described in Tertiary Interaction II?
I would say, yes, but it would be a lot more difficult. Somehow, I can't escape the feeling that the first interaction is meant to stabilize the Y hairpin in a position where the second interaction is much more likely to occur.
Similarly, isn't it conceivable that A-y inserted that way in this lock, contributes to stabilize the X loop in a specific conformation? Possibly, something that would be suitable for the next docking operation, with stack Z perhaps?
This whole affair also made me think of the roles played by the various structural elements in RNA. Considering the loop/stem dichotomy, one can observe that if there would only exist canonical pairs, then RNA would be as boring as DNA. Loops are the essence of the folding of RNA sequences. They give RNA a shape, and ultimately, a function.
There are days where you feel (a lot) more stupid than usual. Guess what I thought that day was for me when I finally inputted the sequence in EteRNA?
Hey, FRABASE sold me that thing for a loop! Look at those starting bases on both sides of the loop, that ain't non-canonical pairs, for (pick your favorite God or Prophet name)'s sake!!
Well, yeah, in a sense, you could say that FRABASE didn't cheat me. Check this out.
GA pair, but it's a "weird" one. Usually, they would occur in a so-called sheared conformation, where the Hoogsteen edge of one base faces the Sugar edge of the other. Here, they are facing "head-on" and both binding by their Watson-Crick edges. This is clearly bound to cause a significant deformation of the helix, possibly such that neighboring bases may have difficulties to actually pair... It may not be obvious in the 3D rendering, but the bases predicted to pair by EteRNA are indeed a little out of range for a proper hydrogen-mediated binding.
There could be another factor explaining the discrepancies. The study indicates a precision of 9.4 Å, which isn't extraordinarily good...
In any case, it seemed to me that the Watson-Crick binding GA was unlikely to occur in a context lacking the tertiary and/or quaternary interactions described above. The sequence would most likely end up forming a much shorter loop, possibly only 11-11, which I judged unacceptable.
So, in the end, no designs to submit.
I'm still undecided whether to thank or to curse him, but undoubtedly, credits have to go to Brourd for subtly inviting me to design a candidate for his lab.
Thanks to Starryjess for proofreading.
And it's only fair:
Molecular graphics and analyses were performed with the UCSF Chimera package. Chimera is developed by the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco (supported by NIGMS P41-GM103311).
If any of this actually increases your taste for spinach, then it is recommended to consult a physician at your earliest convenience.