PROTACs and targeted protein degradation

A classical inhibitor works by occupancy: the effect lasts exactly as long as the molecule sits in the active site, and it needs a continuous concentration above threshold. Two limits follow. The target must have a deep, ligandable pocket — transcription factors and scaffolding proteins do not. And only the function that depends on the occupied site is blocked; the protein’s non-enzymatic scaffolding roles survive.

Targeted degradation changes the question: instead of blocking the protein, remove it.

The ternary complex and the ubiquitin tag

A PROTAC is bifunctional — a target ligand, an E3 ubiquitin ligase ligand, and a linker between them. On its own it inhibits nothing. Its job is to make two proteins that normally never meet do so in a productive geometry.

The assembled ternary complex presents surface lysines on the target for ubiquitin transfer from an E2 enzyme. Successive extension of a Lys48-linked chain is the signal read by the 26S proteasome, which recognises the polyubiquitin, unfolds the substrate and threads it into the proteolytic chamber; the ubiquitin is removed and recycled.

The property that matters is catalysis. Once the target is degraded the PROTAC is released and engages another molecule. The pharmacology is event-driven rather than stoichiometric, so it works at sub-stoichiometric concentrations, and recovery requires resynthesis of the protein — meaning duration of effect is set by protein turnover rather than by the drug’s own pharmacokinetics.

Cooperativity and the hook effect

A good degrader is not the sum of two affinities. What counts is cooperativity: how much binding one partner strengthens binding of the other through new contacts formed between the target and ligase surfaces inside the ternary complex. Linker length and rigidity decide whether that geometry is reachable at all, which makes the linker part of the pharmacophore rather than a spacer. Molecules with mediocre target affinity can be excellent degraders, and the converse holds too.

The same physics produces the characteristic bell-shaped dose-response — the hook effect. In excess PROTAC, target and ligase are each saturated separately, forming two unproductive binary complexes instead of one ternary complex, and degradation falls. It is an unusual case where overdosing mechanistically reduces activity, and it has to be designed around.

Three constraints

Very few usable ligases. The human genome encodes hundreds of E3 ligases, but only a handful have reliable ligands: cereblon within the CRL4 complex (bound by thalidomide analogues), VHL, and a small number of others including MDM2 and the IAP family. Ligase expression sets the tissue selectivity of a degrader — and also explains the main escape route.

Resistance through the machinery, not the target. Unlike inhibitors, where the binding site mutates, degraders lose activity when the ligase itself or components of its complex are lost — CUL4, DDB1, or the neddylation enzymes required to activate cullin-RING ligases. The target sequence is untouched.

Physical chemistry. Degrader molecular weights typically fall between roughly 800 and 1100 Da, well outside the rule of five: many hydrogen-bond donors and acceptors, large polar surface area, high conformational flexibility. Permeability and oral bioavailability are achieved not by obeying rules but through chameleonicity — the molecule folding inside a membrane to bury polar groups in intramolecular hydrogen bonds. This remains the least predictable part of the design.

Cereblon biology also carries a warning: thalidomide’s teratogenicity is attributed to degradation of the transcription factor SALL4, an off-target neosubstrate of the same ligase.