Skip to main content

Chemical Biology

Catching Up With Protein Degradation

Just a short note today – I know that a lot of readers in the Northeast will be snowed out of work today anyway, but there are plenty of others who aren’t! I wanted to mention this short review on targeted protein degradation in J. Med. Chem. (a subject I last wrote about here). It’s a very readable intro to the field, and gets across its promises, uncertainties, and complications. These are still early days, but everyone’s piling into this area (as witness today’s news of the Pfizer/Arvinas deal).

A lot of things have to go right for this technique to work, but the exciting part is that it works as often and as well as it does already. That would have to make you think that as we get a better handle on the various factors that we’ll have even more success in this area, which would give us the ability to target disease in ways that are currently impossible. Those impossibilities come in several forms. You could hit proteins for which we have no real small-molecule inhibitors (since for TPD you just need to find a binder, not necessarily something functional). The downstream effects of protein degradation are also quite different from pharmacological inhibition, as you’d imagine, since you’re yanking all the other protein-protein interactions up by the roots, and that’s also something that’s not been possible to do in this fashion.

The tricky part, though, is that you do need a binder, and there are plenty of interesting proteins for which no believable small molecule binders have been found. My guess (and my hope) is that this situation obtains because so many assays have had functional readouts, rather than just binding-event readouts, and that as we concentrate more on the latter we’ll find more hits that can be turned into degradation “warheads”. But we’ll see about that.


9 comments on “Catching Up With Protein Degradation”

  1. JSR says:

    Seems very tricky to translate the spectrum of what a small molecule binds to, to the spectrum of things degraded when that small molecule is conjugated to cereblon. Nat Gray, Jay Bradner at al have a recent paper that shows this (

    So as this technology moves beyond obvious enzyme targets, for which there are obvious off targets to test for, it will require vigilant attention to what ELSE is being degraded by these degraders. Especially if people get really creative with how they design small molecules to hit “undruggable” targets.

  2. Crank Moody says:

    The biggest challenge for translating this neat concept into drugs will be permeability and oral bioavailability of the molecules. I think PROTACs will end up where other macromolecules that target intracellular proteins are today; interesting in the lab but unlikely to get into the clinic.

    1. 20s says:

      I agree, but this could be a stepping stone to better understand degradomics as a whole. There have been a couple reports of late of how PROTACs are acting more like “molecular glues” based on X-ray and cryo-EM.

      I believe the field is on the cusp of doing better targeted and phenotypic screens with small molecule degraders, like what Celgene is doing with Cereblon E3 and discovering new PPIs with a traditional SM library. I am curious if one day they will find another binding site on Cereblon that doesn’t use the glutarimide ring.

  3. Paul Workman says:

    I think you’re right Derek to highlight that producing targeted protein degraders for hard-to-drug targets does of course still require a binder for the target of interest. I worry when I see this approach described as ‘drugging the undruggable’. It is not drugging the un-ligandable. One needs a binder as the warhead — albeit this is not required to have a direct functional effect on the target. And as you say, it may be easier to discover a ligand by using a binding event readout than with a functional readout assay.

    Not directly related to the above but I think relevant none-the-less, the recent paper in J Med Chem by my colleagues and I at the ICR shows how a heterbifunctional protein degradation probe can be produced successfully against a poorly understood protein with no known catalytic function — in our case the cupin superfamily member and putative transcriptional coregulator pirin — and can then be used to demonstrate intracellular target engagement in absence of downstream molecular pharmacodynamic biomarkers. Our paper also shows how by focusing on linker design and physicochemical properties the effective degrader probe could be achieved in three design iterations. Note however re the first discussion point above that the pirin binding warhead was initially obtained by optimising a hit from a phenotypic screen, with the pirin target identified by affinity matrix chemical proteomics.

  4. Dear Derek, there is a recent publication describing a new technology known as ‘Trim-Away’ that can be very interesting:

    I think it will be interesting to see whether this can translate into therapies.

  5. Radpharmchem says:

    I’m a graduate student (MSc) and new in the field of proteins. I have two questions

    1) is there not still problem of selectivity for the molecules targeting the protein?

    2) if targeting a protein in the brain, there still lies the problem of BBB right?

    1. Derek Lowe says:

      As for #1, definitely. You really want to only degrade the protein you want to degrade, so you’ll want to make a careful choice. As for the second, that’s true as well, and the size of many of these degrader molecules makes that a real factor.

      1. Radpharmchem says:

        Thanks for your reply Derek Lowe.

        This means many of the problems faced with inhibitors still lies in PROTACs.

  6. David Yuan says:

    Dear Derek, I am very interested on DNA-encoded library involving PROTACs. Hope to know whether there are some relating article.

Comments are closed.