Towing Hormone Receptors to the Shredder A Small Molecule Handle for Hijacking the VHL Ligase

This application claims a two-headed drug molecule. One end grabs the estrogen receptor or the androgen receptor, the other end grabs an enzyme called VHL that tags proteins for destruction, and a chemical linker joins the two. The idea is to tow a disease-relevant receptor to the cell’s waste disposal machinery instead of merely blocking it. It was filed jointly by Cambridge Enterprise Limited, GlaxoSmithKline Intellectual Property Development Limited and Yale University, names sixteen inventors led by Craig M. Crews, and is a divisional of Australian application 2024201732.

The Problem

Cells keep their protein levels in check by tagging unwanted proteins with a small protein called ubiquitin, which sends them to the proteasome for destruction. The tagging is done by E3 ubiquitin ligases, and the specification notes that over 600 of them are known in humans. Because each one recognises only certain substrates, they make more attractive drug targets than a general proteasome inhibitor, which hits everything at once.

The difficulty is that E3 ligases work by binding other proteins across large, flat surfaces, and small molecules are poor at wedging into surfaces like that. The specification points to the nutlins as the first small molecule E3 ligase ligands, and says ligands for VHL specifically had been lacking until the inventors’ own work.

The earlier route to the same goal used a peptide. The inventors’ earlier proof-of-concept experiments, which the specification describes, used a short peptide from the protein HIF-1 that VHL naturally recognises. A peptide works in a test tube but makes a poor medicine, because it is large, easily broken down and hard to get into cells. The specification states plainly that a more drug-like replacement was the principal rationale for this work.

What This Invention Does

The title of the application is broad: compounds and methods for the enhanced degradation of targeted proteins and other polypeptides by an E3 ubiquitin ligase. Claim 1 is much narrower, and the article follows the claim. It claims a compound of the form PTM-L-ULM, where PTM is a small molecule that binds an intracellular target protein, L is a linker and ULM is the ubiquitin ligase binding end. In the claim the target is limited to just two proteins, the estrogen receptor and the androgen receptor.

The ULM end is defined by a chemical formula built on a hydroxyproline core, with a free hydroxyl group at one position, an amide-linked aryl group carrying an oxazole, isoxazole, thiazole or isothiazole ring, and a second aryl or heteroaryl arm through which the whole unit is bonded to the linker. The specification explains this is a ligand for VHL, the substrate recognition subunit of the VCB ligase complex. In the cell, VHL normally recognises the oxygen sensing protein HIF-1 once a proline in it has been hydroxylated, and that is what the hydroxyproline core imitates.

Dependent claims narrow the linker. Claim 2 limits the repeat counts in the linker to integers from 1 to 10, and claim 3 limits the linker to a polyethylene glycol chain of 2 to 10 ethylene glycol units. Claim 4 lists specific ULM structures, and later claims name specific PTM structures.

The numbers in the specification come mostly from how the VHL end was built. The inventors ran a fluorescence polarisation test, in which a fluorescent HIF peptide is displaced from VHL by a competing compound, against more than 120 compounds that each carried the central hydroxyproline residue. They then solved co-crystal structures of VHL bound to seven of the best compounds. One of those, compound 15, was resolved at 2.9 angstrom and showed the hydroxyproline sitting in the same pocket as the natural HIF-1 peptide. The two starting fragments were weak, with IC50 values of 117 and 120.1 micromolar, and the specification’s affinity table reports that most of the later compounds were active at about 200 micromolar or lower in the same assay.

Key Features

• Three part molecule. The claimed compound is a target binder, a linker and a VHL binder in a single molecule, so one end finds the receptor and the other recruits the ligase that tags it.
• Hormone receptor targets. Claim 1 restricts the target to the estrogen receptor or the androgen receptor, even though the body of the specification lists many other protein targets, such as BET bromodomain proteins, HDAC and Hsp90 binders.
• Small molecule VHL ligand. The recruiting end is a hydroxyproline based small molecule rather than a HIF peptide, which is the step the specification says makes the approach more drug-like.
• Adjustable linker. The linker can be built from alkyl, ether, thioether, amine, amide or ester units, with ethylene glycol chains of 2 to 10 units claimed specifically.
• Solid phase library chemistry. The specification describes building VHL ligand analogues quickly on a resin, starting from Fmoc protected hydroxyproline attached to a Wang resin and then adding the isoxazole acid and various amines.
• Salts and stereoisomers. Claim 1 covers the compound, its pharmaceutically acceptable salts and its stereoisomers.

Who Is Behind It

The three applicants are an academic university, the technology transfer arm of another, and a pharmaceutical company. Yale University is represented by the laboratory of Craig Crews, who is the lead named inventor. Cambridge Enterprise is the commercialisation arm of the University of Cambridge. GSK is the third applicant. The specification does not say which inventor works for which applicant, so the article does not guess.

The Ciulli laboratory, which now sits at the University of Dundee, records on its own site that a 2009 fellowship took Alessio Ciulli to Yale for a three month visit that began his collaboration with Crews on small molecules for VHL. That is a third party account of the collaboration and not part of the specification.

On priority, the specification says it claims the benefit of United States provisional application 61/585,769, of identical title, filed 12 January 2011. The page headers also carry the numbers WO 2013/106643 and PCT/US2013/021136, and the title page records this as a divisional of Australian application 2024201732. The specification states that the work was made with United States government support under National Institutes of Health grant AI084140. The priority country is therefore the United States.

Why It Matters

Most drugs work by occupying a protein and switching it off for as long as the drug stays bound. A molecule that instead brings the protein to a ubiquitin ligase removes it altogether, and one drug molecule can in principle be reused to destroy many copies of the protein. This design, known as a PROTAC, depends on having a good handle on the ligase side, and a small molecule VHL ligand is exactly that handle.

The choice of estrogen and androgen receptors is pragmatic. Both are long established drug targets in hormone driven disease, so there are well characterised small molecules that bind them and can be attached to a linker. The specification’s disease lists are long and generic, and the claim itself says nothing about a particular disease or about results in cells or animals, so the filing should be read as a claim to the compounds rather than to a proven therapy.

The extracted text of this document shows no cell degradation data for the claimed compounds. The numerical results in the specification concern how well the VHL end binds VHL, not how well the full three part molecule removes a receptor.

Related Concepts

• Targeted protein degradation – the wider field that the claimed compounds belong to.
• Proteolysis targeting chimera – the two headed drug design the claim describes.
• Von Hippel-Lindau tumor suppressor – the protein whose ligase activity the ULM end recruits.
• Estrogen receptor – one of the two target proteins named in claim 1.
• Androgen receptor – the other target protein named in claim 1.
• Hydroxyproline – the amino acid core that the VHL binding end imitates.

---

AU 2026202278 was published in the Australian Official Journal of Patents on 9 April 2026 and is open for public inspection. Patent applications represent inventions that are sought to be protected and do not necessarily reflect commercially available products.