Targeted protein degradation (via PROTACs, molecular glues, etc.) and stabilization (via DUBTACs)

Our bodies use the Ubiquitin-Proteasome System (UPS) to target unwanted proteins (via the addition of chains of a small protein called ubiquitin) to the proteasome (protein-shredder). Much more on the the UPS and how it works “normal-function-wise” in a past post I will link to. But today I want to talk about how we can hijack this system to send specific proteins (maybe ones that are overactive or otherwise causing problems) for degradation or or rescue ubiquitinated proteins from being degraded. The first of those (targeting for degradation) is called Targeted Protein Degradation (TPD) and it uses PROTACs, molecular glues, and other small molecules (drug-like compounds to call over the ubiqutin-adders (ubiquitin ligases). The rescue angle, Targeted Protein Degradation (TPD), uses related molecules called DUBTACs which calls over deubiquitinases to remove ubiquitin tags (kinda like veto-ing the cells’ decision to trash the protein). These strategies are all the rage these days and have a lot of promise, but, unfortunately, also have some challenges. So today, the good, the bad, and the frustrating or targeted protein degradation and stabilization (but mostly degradation).
blog form: bit.ly/tpd_protac_etc
If you need to get caught up first, no problem - here’s that past post on the ubiquitin-proteasome system I mentioned: bit.ly/ubiquitinylation & • The Ubiquitin Proteaso...
Now I’m going to dive in…
In Targeted Protein Degradation (TPD), you induce an interaction between a protein target and an E3 ubiquitin ligase using a small molecule, leading the protein to be ubiquitinated and degraded by the proteasome.
There are two main types of small molecules that can be used for this: PROTACs and molecular glues
- PROTAC stands for Proteolysis-Targeting Chimeras
- these have 2 parts - one part specifically binds to the protein of interest and the other part binds to the ligase
- the ligase-binding part can be hooked up to different protein-targeting parts to target different proteins to the proteasome
- commonly-used ligase-binding parts (and the 2 (out of about 600 human E3 ligases) used most commonly are Cereblon (CRBN) and von Hippel-Lindau (VHL)
- these are just the substrate-binding parts (the substrate receptors) of the ligase complex - like all ubiquitin ligases, they work as part of a larger complex, with additional parts enabling the ubiquitin addition
- in the case of CRBN, it acts as part of the Cullin 4A (CUL4A) E3 ligase complex, with DNA-binding protein 1 (DDB1) and RBX1
- in the case of VHL, it acts as part of the CRL2VHL E3 ligase complex, with Elongin B and C, Cullin 2 (CUL2) and RING box protein 1 (RBX1)
- molecular glue degraders have a single part that binds both the protein of interest and the ligase, promoting an interaction between the two that didn’t exist before or strengthening a weak one
- in the case where a new interaction is formed, we say that the protein target becomes a neosubstrate, with “neo” indicating “new”
- molecular glue degraders are harder to design and are often found “by accident”
- classic examples include thalidomide (which targets CRBN) and analogs of it (such as lenalidomide and pomalidomide)- in fact, thalidomide’s side effects have been attributed to it causing the degradation of
- some of these are referred to as immunomodulatory drugs (IMIDs)
On the flip side of twiddling with protein levels with the help of the ubiquitin-proteasome system is Targeted Protein Stabilization (TPS). This uses Deubiquitinase-Targeting Chimeras (DUBTACs) to call deubiquitinases (DUBs) over to ubiquitinated proteins to remove the ubiquitin tags, rescuing the protein before it would get sent to the proteasome.
- it’s kinda like pulling your retainer out of the trash bag before that bag gets poured into the incinerator? Not a perfect analogy, but hopefully you get the gist
PROTACs and DUBTACs are both examples of heterobifunctional molecules - they have 2 (“bi”) functional parts (parts that “do” stuff”), and those parts and functions are different (“hetero”). In each case, one of the parts, the ligase or DUB-targeting part, is more “generic” in the sense that you can add that lots of different small molecules in order to target lots of different proteins. The other part, however, has to be really specific for the protein of interest. And it has to bind that protein of interest in a spot nearby a lysine amenable to ubiquitination. Oh, and that combination molecule has to successfully make it to and into cells. So it’s a lot harder than it sounds!
Why try?
Why not just go after conventional small molecules, such as just going after that small protein-targeting part? A few reasons.
Finished in comments

[@ThomasTanUltra]

I find this very helpful, thanks for posting! A growing and exciting area for sure with DUBTACs, LYTACs and AUTACs!

[@thebumblingbiochemist]

Great to hear you found it helpful!

[@patrick1073]

Super interesting! I hadn't heard of DUBTACs before :)

[@user-wv7cv3jg7k]

Like your video，thanks very much！
Watching your video, I got several important papers that I didn't notice before.
I am a small molecule drug researcher, and particularly interest in targeted protein degradation.

[@thebumblingbiochemist]

Why not just go after conventional small molecules, such as just going after that small protein-targeting part? A few reasons.
1. you can target sites other than “active sites”
1. traditional drugs bind to the places on proteins where the protein “does stuff” like binding to other molecules and, in the case of enzymes, catalyzing (speeding up/facilitating) reactions, thereby preventing the proteins from doing that stuff
2. these active sites (especially ones where reactions are catalyzed) are often similar between different enzymes because similar enzymes use the same mechanisms to carry out the “same” reactions (e.g. phosphorylation, hydrolysis) on different substrates (starting molecules)
1. this makes it hard to target individual enzymes specifically - instead you often end up accidentally targeting other important molecules causing off-target effects
3. not all proteins have traditional “active sites” to target - and some active sites have been considered “undraggable” because they don’t have much in the way of binding opportunities (shallow pockets, etc.)
1. with these heterobifunctional molecules you can use “any” binding site on the protein, so if the active site isn’t ideal, find another site that’s better!
2. you don’t need to use as much
1. these molecules act substoichiometrically, meaning that a single copy of the drug can lead to the degradation or stabilization of multiple copies of the target
1. the molecule doesn’t have to stick around once it recruits the helpers - this is as opposed to traditional small molecules that basically just have to sit there doing nothing to prevent the protein from doing anything (or get it to do something, but such activating drugs are much more rare)
What can go wrong?
It can be hard to design a heterobifunctional compound. It’s more difficult than just chemically stitching together two things which you know bind to your 2 targets. If you were to simply “paste” them together, they probably wouldn’t bind those 2 targets anymore because there’s not enough room. Picture that dog spaghetti scene from Lady and the Tramp. If the spaghetti’s super duper short there’s not enough room even for a kiss! (because the teeth, lips, etc. take up space). But, if you lengthen that spaghetti, now you can get a romantic love scene. Similarly, if you add a linker between the molecules, you can allow them to still bind their targets (hopefully!). Optimizing the linker length can be a challenge though, for a variety of reasons outside the scope of this post. And as you go longer, you make bigger molecules that will likely have a harder time getting into cells.
Even if the drugs bind, and even if they form a ternary complex (protein, drug, and ligase (or DUB)), the complex might not be able to actually do what you want it to. The ubiquitination site has to be in an accessible location to the active site of the ligase or DUB.
Here’s a good review article: Cowan, A. D., & Ciulli, A. (2022). Driving E3 Ligase Substrate Specificity for Targeted Protein Degradation: Lessons from Nature and the Laboratory. Annual review of biochemistry, 91, 295-319. doi.org/10.1146/annurev-biochem-032620-104421
Another increasingly-recognized challenge is that the activity of PROTACs can vary depending on the sub-cellular localization of the target, in combination with the target itself. See this recent article for more:
Simpson, L. M., Glennie, L., Brewer, A., Zhao, J. F., Crooks, J., Shpiro, N., & Sapkota, G. P. (2022). Target protein localization and its impact on PROTAC-mediated degradation. Cell chemical biology, 29(10), 1482-1504.e7. doi.org/10.1016/j.chembiol.2022.08.004
Although it’s been proposed to be more resistant to resistance development than traditional drugs, like basically all anti-cancer drugs, resistance can arise since cancer cells can randomly acquire mutations and select for advantageous one so rapidly due to their rapid growth rate. This resistance can come from mutations in the E3 ligase being recruited and/or other components of the UPS.
How can it be improved?
One area of future development is trying to restrict the drugs’ activity to specific tissues or cell types. Some drugs go about this by targeting where the drug gets trafficked to in the body (taking advantage of cell-type specific surface proteins, etc.) but this can be challenging. Since there are over 600 E3 ligases, and some of them are expressed differentially in different cell types, there’s another way to restrict the drugs’ action - let the drugs go everywhere in the body, but only “let” them do stuff in the cells you want them to do stuff in. You could (at least theoretically) do this by targeting ligases that are only (or at least predominantly) expressed in the cells of interest. Unfortunately, not much work has been done yet on targeting E3 ligases other than the classical 2, CRBN and VHL). But it’s definitely an area to watch!
Here’s a cool preprint I came across on this and related topics:
Proteome-scale induced proximity screens reveal highly potent protein degraders and stabilizers. Juline Poirson, Akashdeep Dhillon, Hanna Cho, Mandy Hiu Yi Lam, Nader Alerasool, Jessica Lacoste, Lamisa Mizan, Mikko Taipale. bioRxiv 2022.08.15.503206; doi: doi.org/10.1101/2022.08.15.503206
There’s a reason those two are often used though - they work better than other things that have been tried (at least those that have been tried so far with the targets so far chosen). For example, this group explored using a different E3 ligase, KEAP1, and they didn’t have much luck - they could get it to degrade a couple things, but couldn’t get it to degrade other things they could get CRBN and VHL to degrade…
Du, G., Jiang, J., Henning, N. J., Safaee, N., Koide, E., Nowak, R. P., Donovan, K. A., Yoon, H., You, I., Yue, H., Eleuteri, N. A., He, Z., Li, Z., Huang, H. T., Che, J., Nabet, B., Zhang, T., Fischer, E. S., & Gray, N. S. (2022). Exploring the target scope of KEAP1 E3 ligase-based PROTACs. Cell chemical biology, 29(10), 1470-1481.e31. doi.org/10.1016/j.chembiol.2022.08.003
Other related areas to watch:
Since the development of PROTACs (and then DUBTACs), others have hopped on to the heterobifunctional molecule train to target other means of cellular waste disposal - these include autophagy-targeting chimeras (AUTACs) and lysosome-targeting chimeras (LYTACs). LYTACs have the benefit of being able to target & degrade membrane-bound and extracellular proteins. Oh - and, by the way, they were actually developed by the lab of Carolyn Bertozzi, who you might remember, just won a Nobel Prize in Chemistry for her work on bioorthogonal chemistry!
Banik, S. M., Pedram, K., Wisnovsky, S., Ahn, G., Riley, N. M., & Bertozzi, C. R. (2020). Lysosome-targeting chimaeras for degradation of extracellular proteins. Nature, 584(7820), 291-297. doi.org/10.1038/s41586-020-2545-9
more on Carolyn Bertozzi’s other work, on bioorthoganal chemistry: blog: bit.ly/bioorthogonal_prize ; CZcams: czcams.com/video/LrwzFRY7gsk/video.html
more on DUBTACs: Henning, N. J., Boike, L., Spradlin, J. N., Ward, C. C., Liu, G., Zhang, E., Belcher, B. P., Brittain, S. M., Hesse, M. J., Dovala, D., McGregor, L. M., Valdez Misiolek, R., Plasschaert, L. W., Rowlands, D. J., Wang, F., Frank, A. O., Fuller, D., Estes, A. R., Randal, K. L., Panidapu, A., … Nomura, D. K. (2022). Deubiquitinase-targeting chimeras for targeted protein stabilization. Nature chemical biology, 18(4), 412-421. doi.org/10.1038/s41589-022-00971-2
A recent paper found that C-terminal cyclic imides can form as degrees that recruit CRBN. These amides can form spontaneously during asparagine (Asn) or glutamine (Gln) deamidation or when the side chain attacks the protein backbone leaving to protein cleavage and a new C-terminus with a cyclized end.
Ichikawa, S., Flaxman, H. A., Xu, W., Vallavoju, N., Lloyd, H. C., Wang, B., Shen, D., Pratt, M. R., & Woo, C. M. (2022). The E3 ligase adapter cereblon targets the C-terminal cyclic imide degron. Nature, 10.1038/s41586-022-05333-5. Advance online publication. doi.org/10.1038/s41586-022-05333-5
more on molecular glues:
Dong, G., Ding, Y., He, S., & Sheng, C. (2021). Molecular Glues for Targeted Protein Degradation: From Serendipity to Rational Discovery. Journal of medicinal chemistry, 64(15), 10606-10620. doi.org/10.1021/acs.jmedchem.1c00895
    
more about all sorts of things: #365DaysOfScience All (with topics listed) 👉 bit.ly/2OllAB0 or search blog: thebumblingbiochemist.com

[@merlin4025]

This is gem!

[@josemanuelnegroe9992]

great video, thank you

[@thebumblingbiochemist]

Thanks so much!