PROTACs and oral bioavailability — what bRo5 chemistry actually requires

Lipinski’s Rule of Five was empirical: orally absorbed drugs tend to weigh less than 500 Da, have fewer than five hydrogen-bond donors, fewer than ten acceptors, and a cLogP under 5. PROTACs violate every one of those numbers and the early ones in the clinic were IV-dosed because of it. Then vepdegestrant became the first PROTAC to win a Phase 3 with oral dosing and the question stopped being “can a bRo5 PROTAC be oral” and started being “which physicochemical parameters actually matter when you are already past Lipinski.” The answer is more useful than the original rule.

Why PROTACs sit outside Ro5

A PROTAC is a heterobifunctional molecule: a ligand for the protein of interest (POI), a ligand for an E3 ubiquitin ligase (typically VHL or CRBN), and a linker connecting them. The molecule does not occupy a binding site so much as bridge two proteins, recruiting the E3 ligase to ubiquitinate the POI, which is then degraded by the proteasome. The molecule itself is catalytic — it dissociates and recycles after each degradation event.

The chemistry of doing that with small organic molecules ends up between 700 and 1100 Da, with a topological polar surface area (TPSA) in the 150-250 range and hydrogen-bond counts well above Lipinski limits. The molecule is not a Lipinski drug. The question is whether it is in the “beyond Rule of Five” (bRo5) space where cyclic peptides and macrocycles already showed oral absorption is possible, or in the genuinely non-oral space.

The empirical bRo5 window for PROTACs

Pike et al. (J Med Chem 2023) and the Arvinas team (independently, in the AstraZeneca-collaborative work) looked at hundreds of PROTACs in rat oral absorption assays and found a usable window. The headline numbers, for CRBN-based PROTACs with bioavailability above 20%:

• Molecular weight 700-900 Da — CRBN PROTACs can absorb well in this band; VHL PROTACs typically run heavier (above 1000 Da) and get poor oral exposure as a result.
• cLogP 3-6 — higher than Ro5 allows, but the lipophilicity is needed to balance the polar surface required for two ligand pharmacophores plus a linker.
• TPSA up to 200 — well past the 140 ceiling that Veber’s rules suggested for oral absorption. PROTACs break that.
• Exposed HBDs (eHBD) ≤ 2 — this is the parameter the Arvinas and AstraZeneca papers both isolated as the strongest predictor. The number of hydrogen-bond donors that are solvent-exposed (not engaged in an intramolecular interaction) is what kills membrane permeability. A molecule with 5 HBDs but only 2 unsatisfied can still cross the membrane; one with 3 HBDs all exposed cannot.

The eHBD insight is the actionable design rule that came out of this. It explains why two molecules with identical Lipinski descriptors can have a 10-fold gap in bioavailability — the difference is whether the linker conformation buries the donors.

Vepdegestrant: the clinical proof

Vepdegestrant (ARV-471, Arvinas/Pfizer) is an oral PROTAC degrader of the estrogen receptor. It launched as a VHL-based molecule and the chemistry team switched to a CRBN ligand, which dropped molecular weight by roughly 200 Da and brought the polar surface area into the absorption window. The Phase 3 VERITAC-2 study in ER+/HER2- metastatic breast cancer with ESR1 mutations reported a statistically significant progression-free survival benefit over fulvestrant — the previous standard injectable degrader. Arvinas and Pfizer submitted the NDA in June 2025; the molecule earned Fast Track designation in 2024.

The clinical milestone matters because every previous PROTAC that reached patients was IV-dosed. Vepdegestrant is the existence proof that a 720 Da heterobifunctional degrader with TPSA around 190 can be a daily oral tablet — provided the eHBD count is constrained and the linker geometry encourages intramolecular hydrogen bonding. That blueprint is now being applied across the PROTAC pipeline.

What this means for docking workflows

Two practical implications. First, conventional docking against a single protein under-describes a PROTAC’s behavior — the relevant geometry is the ternary complex of POI, PROTAC, and E3 ligase. Tools like PRosettaC, MOE’s PROTAC module, and Rosetta-based ternary docking exist for that problem; standard Vina-class docking is not built for it. Second, ADMET prediction at the bRo5 boundary is harder than for Ro5 compounds because the training distributions are sparser. Models from admet-ai and ADMETLab were trained primarily on small molecules and will systematically under-predict the oral absorption of well-designed PROTACs and over-predict it for poorly designed ones.

The pragmatic workflow today is to dock the POI warhead separately to confirm it engages the target, dock the E3 ligand to confirm it engages VHL or CRBN, and run ternary modeling in a dedicated tool. The eHBD count, MW, and TPSA you compute from the resulting structure are the descriptors that should drive linker iteration.

Try the warhead docking yourself

While Liganx does not currently do ternary-complex modeling, warhead optimization on the POI side is exactly what Liganx is built for. Open Studio and dock your candidate POI-side ligand against the target. The ADMET panel flags molecular weight, logP, and TPSA explicitly on every result row, so you can see at a glance whether the warhead alone is consuming so much of the bRo5 budget that the final PROTAC will not be orally bioavailable. Aim to keep the warhead under 350 Da and below cLogP 4 if you want headroom for the linker plus E3 ligand.

Liganx is molecular docking online and free in the browser, useful for the target-side optimization step even when the final molecule is bRo5. Using molecular docking on the warhead before committing to a linker is the cheapest place to fix physicochemical drift.

Primary sources

• Edmondson SD, Yang B, Fallan C. Phenotypic drug discovery for protein-protein degraders: physicochemical property determinants of oral absorption for PROTAC protein degraders. J Med Chem 66, 8810-8839 (2023). doi:10.1021/acs.jmedchem.3c00740
• Pike A, Williamson B, Harlfinger S, Martin S, McGinnity DF. Optimising proteolysis-targeting chimeras (PROTACs) for oral drug delivery: a drug metabolism and pharmacokinetics perspective. Drug Discov Today 25, 1793-1800 (2020). doi:10.1016/j.drudis.2020.07.013
• Hamilton EP, et al. Vepdegestrant, a PROTAC estrogen receptor degrader, in advanced ER-positive HER2-negative breast cancer. Nat Med 30, 3289-3297 (2024). doi:10.1038/s41591-024-03245-7
• Snyder LB, et al. NDA submission of vepdegestrant (ARV-471) to U.S. FDA: the beginning of a new era of PROTAC degraders. J Med Chem (2025). doi:10.1021/acs.jmedchem.5c01818