If anyone’s wondering what I’ve been up to these last 5 years… dozens of mutant proteins, hundreds of slot-blots, countless scintillation counting kinase assays, 262 HDX-MS analyzed peptides, & lots of trouble-shooting later I can finally tell you!

Here’s a Tweetstorm all about my preprint, a preprint you can find here: https://www.biorxiv.org/content/10.1101/2022.01.06.475261v1

In RNA interference (RNAi), an Argonaute (Ago) protein uses a small RNA (e.g. miRNA or siRNA) as a guide to find, bind, and induce the repression of target mRNAs containing regions of complementarity. 2/

But then what? You want to reuse that Ago/guide complex (aka RISC), right? Which means it’s gotta let go! Our work might help explain how cells can give it a nudge 🙂 3/

We show target-binding triggers hierarchical phosphorylation by CK1α of a region of Ago called the eukaryotic insertion (EI) & the negative charge added adds up & clashes with the negative charge of the target backbone, leading to target release, freeing RISC to repress 4/

(Since the guide is still tightly bound, once Ago gets dephosphorylated, this RISC can do it again). So, you want the deets? I mean, it is 5 years worth of work… Let’s start with the target-triggering… 5/

I had a bit of a panic early on in my project when I tried to get CK1α to phosphorylate RNA-free Ago and it wouldn’t… It wouldn’t phosphorylate guide-bound either. But I added the target and boom! EI Phosphorylation! But why? What was magic about it? 6/

We saw we needed a couple features of the target to trigger phosphorylation. To narrow things down we tried to figure out the minimal requirements, starting by shortening the target. We found that once we went below 14nt, there was almost no phosphorylation 7/

Why could this be? Could it have to do with supplemental pairing? The main way RISC recognizes targets is through the 5’ seed sequence, with variable amounts of 3’ complementarity helping provide greater specificity and affinity between miRNA family members. 8/

We found that strong triggering of phosphorylation requires pairing in the 3’ supplemental region & suggest phosphorylation could allow RISC to take advantage of the specificity & affinity provided by supplemental pairing without having to worry about getting stuck. 9/

“Getting stuck” could be a concern for such seed+supplemental targets because they lack the central region pairing required for slicing, a process in which Ago cleaves the target and which promotes release. 10/

Fully-complementary targets (siRNA-like) did not trigger phosphorylation, and, combined with the evolutionary pattern of the EI conservation among only miRNA-handling Agos, this suggests this phosphorylation is a miRNA-specialized mechanism. 11/

Most of our work was carried out on human Ago2 (our main Ago protein) but we also found that other human Agos similarly got phosphorylated upon target-binding, suggesting this mechanism could be wide-spread. 12/

But *how* does target-binding trigger phosphorylation? We wondered if the EI might be undergoing conformational changes upon target-binding, but were a bit stuck crystallography-wise because the EI is unresolved in all Ago structures. So we turned to HDX-MS. 13/

HDX-MS (hydrogen-deuterium exchange mass spectrometry) lets you analyze differences in secondary structure and solvent accessibility at the peptide level by measuring the extent of exchange of amide hydrogens for deuterium. 14/

Strong secondary structure and/or solvent inaccessibility can protect regions for exchange, whereas solvent-exposed, flexible regions will show high uptake. 15/

With Betsy Komives and Steve Silletti at UC San Diego, we performed HDX-MS on Ago in the RNA-free, guide-bound, & guide+target-bound states to look for any potential differences in the EI. And?… 16/

Nope. We didn’t find any evidence of target-binding leading to changes we could detect with this method – and we saw a high level of exchange in all of their EI’s, suggesting the region likely has little if any defined secondary structure. 17/

Speaking of high exchange… when we looked at the RNA-free Ago, we saw very high levels of exchange, consistent with limited proteolysis results showing that RNA-free Ago is highly flexible and gets locked into place by guide-binding. 18/

Why’s this relevant to our story about target-binding? If target-binding trigger phosphorylation due to a conformational change outside of the EI making the EI more accessible for CK1α, we’d expect CK1α would have an easy time phosphorylating RNA-free Ago. But it doesn’t! 19/

What if CK1α actually needs that target RNA? What if the RNA has a larger role to play in the story? Could it be directly interacting with CK1α? Maybe even non-cannonically-priming it? Let me step back and explain a bit… 20/

CK1α prefers to phosphorylate Ser or Thr 3 residues downstream of a previously phosphorylated “priming” site. This allows it to carry out hierarchical phosphorylation. 21/

And the kicker I haven’t told you about yet is that we found that CK1α can phosphorylate Ago in the absence of priming phosphorylation (more below)! It doesn’t even need acidic upstream residues to act as an alternative priming source. So maybe the RNA backbone can do it?! 22/

We found CK1α could phosphorylate S828 (and only S828) in the absence of priming. 23/

And it doesn’t stop there. Once you phosphorylate 828 you have a priming source for additional phosphorylations, so we see hierarchical phosphorylation occurring. (which, with 5 potential sites was complicated to tease apart…) 24/

And the added negative charges start clashing with the negatively-charged target backbone, electrostatically-repulsing the target, leading to its release. 25/

All in all, we have this model whereby target-binding triggers hierarchical phosphorylation of the EI (proposedly after repression), which promotes target release, freeing RISC (once deposphorylated) to repress more targets 26/

We propose that this helping balance RISC efficiency and efficacy in a world where potential targets > RISC /end (and thanks for reading!)


Originally tweeted by Brianna Bibel (@biochem_bri) on January 7, 2022.

Leave a Reply

Your email address will not be published.