Skip to Content

Biological News

Is Selective Ribosome Stalling Possible? Apparently So

PCSK9 is a drug target that’s famous in several directions. If you’re interesting in human genetics, it’s famous as an example of a “human knockout” – people with nonfunctional PCSK9, and there are a handful, have extraordinarily low levels of LDL, a finding that immediately got drug companies interested in finding inhibitors. If you’re interested in the business side of the industry, then PCSK9 is famous because there are competing antibodies from Amgen and Regeneron/Sanofi fighting it out in the market, with a Pfizer one coming along shortly. Similarly, if you’re into regulatory issues and outcomes, this is a big area as well. LDL lowering is something that these therapies certainly can deliver – but does that translate into real-world endpoints in mortality and morbidity? Since the progress of cardiovascular disease is slow, no one’s quite sure yet. Both players in the field are gathering data, in the hopes that their products will see much wider use once their impact on patient outcomes clears up – it’s safe to say that uptake of both drugs has been slower than some people had hoped, for just this reason.

pf-compoundNow we’re back to another biochemical reason to find this target interesting. A joint paper from a team at Pfizer and one at Berkeley (the Cate group) is out on an interesting and unusual mechanism. They report a small molecule, PF-06446846, inhibits PCSK9 expression by directly acting on the ribosome, stalling it at a particular codon as the protein is being assembled. Ribosome stalling is a topic that’s been around for many years – there are several antibiotics (like erthythromycin) that use it as their mechanism of action (it can also be part of normal gene regulation in some bacteria). The molecule itself (at right) is not particularly small (molecular weight of 433), but it’s certainly not a gigantic weirdo thing, either. There’s nothing about it that would make you think “Yeah, that’s going to do something unexpected”; it’s an assembly of perfectly reasonable med-chem pieces. You never know!

This series of compounds was discovered by phenotypic screening, and I’ll bet that the mechanism came as a surprise. It certainly surprised me – you wouldn’t think that monkey-wrenching the ribosome would be (or even could be) a very specific process, which is surely why it shows up in the kill-or-be-killed world of antibiotics. A number of control experiments ruled out general effects on protein synthesis, though, or mechanisms targeting messenger RNA. Instead, it looked as if the drug’s effect depended on the specific amino acid sequence of PCSK9 itself, which suggests something to do with the ribosome’s “tunnel” region, which the protein chain moves through during the synthesis process.

There are peptides known that do similar things, but (to the best of my knowledge) this is the first small molecule that does it in such a protein-specific manner. The details of the mechanism are still fuzzy, because although there are a few other proteins whose expression is also altered, they don’t have much in the way of common features – you can’t just predict by the primary sequence what’s going to be affected. There’s something subtle going on in the ribosomal exit tunnel that’s going to require a lot of work to figure out.

But that work could well be worthwhile. Being able to do this specifically, and to order, would be a major advance. There’s no guarantee that such a broad application of this concept is possible – in fact, there are many reasons to think that it might well not be – but this example alone is more than most people would have thought likely. Affecting transcription and translation through small molecules is a tricky business, and there have been a lot of failures in the area over the years. The biology is ridiculously complex, when viewed through a small-molecule lens, and a lot of drug discovery efforts have disappeared into those swamps. Is this a patch of dry ground, or not?

14 comments on “Is Selective Ribosome Stalling Possible? Apparently So”

  1. YoloPeptide says:


  2. Glen Weaver says:

    There is no dry land in that swamp. Though a resting alligator may appear so. It may slowly move away, taking you to an interesting area. It may sink, leaving you to swim or drown.It may grab you to stash in the mud for a later dinner. It is never safe, dry land.

  3. Mol Biologist says:

    IMO ti has nothing to do with Ribosome Stalling. PF-06446846 very likely activates stress response of Unfolded protein response UPS with diminished levels of major endoplasmic reticulum chaperones GRP78 (BiP) and GRP94.

    1. Astor says:

      A general cell stressor should have a global impact on translation in the cell based on the PERK->ei2F-p pathway slowing down all translation. From the paper:
      “A number of control experiments ruled out general effects on protein synthesis, though, or mechanisms targeting messenger RNA. Instead, it looked as if the drug’s effect depended on the specific amino acid sequence of PCSK9 itself.”
      This seems to argue against general cell stress/UPR-type mechanisms I would think.

  4. Barry says:

    If this mechanism of action stands up to further scrutiny, it’s a whole new field for small-molecule intervention, effectively blocking mammalian transcription factors without having to get into the nucleus. Mammalian transcription factors have proven to be very, very hard targets, but they’re prominent in a bunch of disease states, cancer prominent among them.
    Erythromycin and related macrolide antibiotics–if I understand it correctly–work by blocking bacterial ribosomes in the presence of mammalian ribosomes. That’s different from discriminating with regard to the mRNA sequence

    1. ScientistSailor says:

      Macrolides do inhibit translation by binding to the nascent peptide in the tunnel. Different macrolides inhibit different peptide sequences. This isn’t new. What’s new here is seeing such selectivity.

      1. Barry says:

        macrolide antibiotics–as far as I know–bind to the bacterial ribosome, and make no contacts to the mRNA or the produced peptide. Although it seems some peptide sequences can sometimes dislodge the antibiotic

  5. Chrispy says:

    Specificity would require binding an RNA or DNA sequence of ten bases or more.(?) This would require something much larger than a typical small molecule drug. And then you have the delivery issues of the RNA folks. No, thanks. Even the macrolides are big for simple monkey wrenches. I hope they prove me wrong, but this approach seems fundamentally flawed.

  6. tangent says:

    “A number of control experiments ruled out general effects on protein synthesis, though, or mechanisms targeting messenger RNA. Instead, it looked as if the drug’s effect depended on the specific amino acid sequence of PCSK9 itself”

    They ruled out fully general effects, okay, but how about a little bit general? The effect may be pretty specific, and hit only 0.01% of proteins with a suitable sequence. That’s a lot.

    I am skeptical that it’s so specific as to hit this and nothing else. Look at it this way, how many ‘siblings’ does the molecule have, who also have the stalling effect but hit on different sequence strings? You need billions of those, for it the pool to likely offer one that hits just what you’re targeting. If the molecule were a big hairy macrolide, okay, maybe, but with a molecule as small as this, you’d have to get incredibly lucky. Bayes’ Law suggests it’s more likely an experiment tricked you.

    1. tangent says:

      The example from erythromycin resistance is actually backwards: it’s the bacterial answer to “given this macrolide drug, can we find an oligopeptide sequence that will interact with it?” It’s not an answer to “given an oligopeptide sequence, can we find a drug molecule that interacts with it?” If there are more sequences than there are drugs, then by the numbers the bacteria’s problem is easier than the drug-finder’s.

      (Also, the drug-finder needs specificity, but the bacteria don’t care much — if the sequence interacts similarly with some “off-target” molecule, who cares, that molecule probably wasn’t desired hanging around in the ribosome either!)

  7. pcsk9scientist says:

    The idea is interesting, but the results are questionable. To be fair they used rats, but with mice (and rats too), plasma cholesterol on a standard chow diet is primarily HDL and PCSK9 inhibition has very minimal effects unless it is fully eliminated in mice. This paper is reporting a 30% total cholesterol reduction, but even at the highest dose PCSK9 levels are going down by maybe around 60%. PCSK9 knockout mice have about a 40% reduction in total cholesterol, liver specific PCSK9 knockout mice have about 30% reduction in total cholesterol, both of these knockouts essentially eliminate all PCSK9 in the blood and HDL is lowered in both groups. But HDL isn’t going down with this compound, or at least not significantly, and yet total cholesterol is going down. PCSK9+/- mice have a 70% reduction in circulating PCSK9, but only a 1.3-fold induction in LDLR protein in the liver (compared to PCSK9-/-, which have a 3.2-fold increase in liver LDLR with a 100% reduction in circulating PCSK9, suggesting you need a strong PCSK9 reduction to get big LDLR effects in rodents on chow diet). There is likely something else going on with this Pfizer compound at the high dose (50mg/kg), which seems to be the only dose really responding to drive down total cholesterol; LDL seems to be going down with the 15 mg/kg dose, but the reduction in PCSK9 is about 30% at that dose, which is similar to the reduction in serum PCSK9 seen in Aplp2-/- mice that don’t have any increase in liver LDLR levels. The data don’t add up for a specific PCSK9 inhibition effect.

    1. Barry says:

      how shall we understand a reduction >50% in circulating PCSK9 in the heterozygous mouse? Is there a saturable non-circulating compartment?

      1. pcsk9scientist says:

        I’m not sure, but rodent pcsk9 reduction could lead to modest increases in LDLR (via less degradation) that then leads to more cholesterol uptake that slightly inhibits SREBP transcriptional activity by SREBP rentention when bound to sterol bound SCAP (SREBP processesing inhibition). Lower SREBP activity would lead to less PCSK9 production but also lower LDLR transcription, which may be why the LDLR increase is modest with the remaining (albeit much lower) circulating PCSK9 still able to bind LDLR and lead to lysosomal degradation (the complete PCSK9 knockout wouldn’t have this latter issue which could be why the LDLR levels are much higher).

  8. Arne says:

    Who cares? The better question is, if PIs stopped treating their students like garbage, would science progress like it did circa 1985?

Comments are closed.