Switching a Blood Vessel Receptor Back On An Antibody That Activates Tie2 Instead of Blocking Its Antagonist

This application covers antibodies that bind a defined region of the Tie2 receptor on blood vessel cells and switch it on, together with the genetic material to make them and pharmaceutical compositions containing them. The target is angiogenic disease: tumours, diabetic eye disease, sepsis and a long list of conditions in which blood vessels are structurally and functionally abnormal. The applicants are Korea’s Institute for Basic Science and KAIST.

The Problem

Blood vessels are not merely plumbing that happens to be present in a diseased tissue; in many diseases they are part of the pathology. Tumour vessels are leaky and disorganised, which starves the tumour core of oxygen, drives progression and metastasis, and simultaneously prevents anticancer drugs from reaching the cells they are meant to kill. The same defect appears in diabetic macular oedema, wet macular degeneration, viral infection and the acute inflammatory collapse of sepsis. A drug that could normalise pathological vessels rather than simply destroy them would apply across all of it.

The switch that controls vessel stability is the Tie2 receptor, encoded by TEK and sitting on the surface of vascular endothelial cells. Two members of the angiopoietin family compete at it. Angiopoietin-1 binds Tie2, phosphorylates and activates it, and vessels become stable and tight. Angiopoietin-2 binds the same receptor but acts as an antagonist, inactivating Tie2 so that vessels destabilise and leak. Ang2 is elevated in the blood of cancer patients and in ocular, infectious and inflammatory disease, which made it an obvious drug target.

Obvious, but not productive. The specification catalogues a decade of disappointment with the neutralise-Ang2 strategy. Numerous anti-Ang2 antibodies blocked Ang2 binding and showed anti-angiogenic activity in preclinical models, yet clinical efficacy was insufficient: an Amgen Phase 3 trial in ovarian cancer produced an insignificant result, and the Ang2 antibody nesvacumab failed to improve on anti-VEGF therapy in a Phase 2 combination study in ocular patients.

Direct activation of Tie2 was tried instead, with its own problems. Engineered Ang1 variants such as COMP-Ang1 and the peptide mimetic Vasculotide activated the receptor but had very short half lives and unstable physicochemical properties. The small molecule AKB-9778 raised Tie2 activity indirectly by inhibiting the phosphatase VE-PTP, at the cost of activating other receptors as well. Agonistic Tie2 antibodies had been described, but the field still lacked an antibody with the affinity, stability and specificity to work as a therapeutic.

What This Invention Does

The inventors set out to build antibodies that bind Tie2 directly and turn it on. What distinguishes the result is where they bind. Claim 1 covers an anti-Tie2 antibody, or an antigen binding fragment of one, that binds the Ig3-FNIII (1-3) domain of the receptor, with a dependent claim narrowing the site further to amino acids 633 to 644 and amino acids 713 to 726 of the Tie2 sequence. That is not the ligand binding pocket where the angiopoietins compete; it is a separate region of the extracellular stalk, and binding there triggers phosphorylation and activation of the receptor.

The consequence matters clinically. An agent that works by blocking Ang2 is fighting a concentration battle, and its effect depends on how much Ang2 is present and on Ang2’s confusing dual behaviour, since the same protein acts as an agonist in lymphatic vessel formation and maintenance. An antibody that activates Tie2 at a site the angiopoietins do not occupy drives the receptor towards its stabilising state regardless of the ligand balance around it. It is an agonist monoclonal antibody, a considerably harder thing to engineer than a blocker, because activation usually depends on holding two receptor molecules in the correct geometry rather than simply covering a binding site.

Beneath the epitope claim, the application also claims specific antibodies. Five sets of heavy and light chain complementarity determining regions are recited by sequence, along with a longer list of paired heavy and light chain variable regions, which is the standard way of defining named clones without naming them. The remaining claims build out the manufacturing chain: nucleic acids encoding the antibodies, an expression vector containing the nucleic acid, a cell transformed with the vector, and a production method consisting of culturing that cell and recovering the antibody from it.

The therapeutic claims are broad. A composition containing the antibody as its active ingredient is claimed for preventing or treating angiogenic disease, and the definition of angiogenic disease runs from cancer and metastasis through diabetic retinopathy, retinopathy of prematurity, corneal graft rejection, macular degeneration and neovascular glaucoma to psoriasis, rheumatoid arthritis, atherosclerosis, diabetic nephropathy, Crohn’s disease and neurodegenerative disease. One of the classifications on the application, A61K 45/06, covers mixtures of active ingredients, consistent with the stated aim of co-administering the antibody with existing therapies rather than replacing them.

Key Features

• Epitope defined claim. The primary claim is written around the binding site, the Ig3-FNIII (1-3) domain of Tie2, rather than around a single antibody sequence.
• Agonist rather than antagonist. The antibody activates Tie2 by inducing phosphorylation, instead of neutralising the Ang2 that inactivates it.
• Binding away from the ligand pocket. The narrowed epitope, amino acids 633 to 644 and 713 to 726, sits outside the region where angiopoietins compete, so activity does not depend on the local Ang1 to Ang2 ratio.
• Five defined clone families. Distinct heavy and light chain CDR sets and variable region pairs are recited by sequence, giving specific fallback positions beneath the broad claim.
• Full production chain claimed. Nucleic acids, expression vector, transformed cell and a culture and recovery method are claimed alongside the antibody itself.
• Combination ready composition. The pharmaceutical claims extend to co-administration with other agents for angiogenic disease across oncology, ophthalmology and inflammation.

Who Is Behind It

The Institute for Basic Science is a Korean government funded research institute headquartered in Daejeon, established in 2011 to run long horizon basic research through a network of about thirty centres. KAIST, also in Daejeon, is Korea’s leading science and technology university. The two institutions co-own this filing because the work sits across both: the IBS Center for Vascular Research is hosted on the KAIST campus.

The lead inventor is Gou Young Koh, founding director of the IBS Center for Vascular Research and a distinguished professor at the KAIST Graduate School of Medical Science and Engineering. He is one of the more significant figures in vascular biology, having spent decades on the growth factors that regulate the microvasculature, and his group generated several of the angiogenesis modulators referenced in this field, including COMP-Ang1 and the Ang2 binding and Tie2 activating antibody ABTAA. The co-inventors, Jeomil Bae and Jaeryung Kim, are researchers in the same group.

The application is a divisional of Australian application 2019283520, the national phase entry of PCT international application PCT/KR2019/006820 filed on 5 June 2019. That PCT claims priority from a United States provisional application filed on 7 June 2018 and a South Korean application filed on 5 June 2019.

Why It Matters

The commercial case for the angiopoietin-Tie axis stopped being theoretical during the life of this family. Faricimab, a bispecific antibody that binds both Ang2 and VEGF-A, was approved in 2022 for diabetic macular oedema and neovascular macular degeneration and became a major ophthalmology product, demonstrating that intervening at this pathway alongside anti-VEGF therapy produces real clinical benefit. That validates the target while leaving open the question this application addresses, which is whether activating Tie2 directly is better than removing its antagonist.

Technically, the interesting move is the shift from blocking to switching on. The great majority of therapeutic antibodies work by occlusion, because covering a binding site is a tractable engineering problem. Agonist antibodies have to reproduce a signalling event, which depends on the geometry of receptor engagement rather than on affinity alone, and the field’s growing ability to design them opens targets that antagonists cannot reach. Choosing an epitope in the Ig3-FNIII stalk rather than the ligand pocket is a deliberate expression of that approach.

The filing strategy is characteristic of institutional science that intends to license. Broad epitope level claims backed by specific sequences give a partner room to develop a clinical candidate while retaining coverage against competitors who arrive at a different antibody against the same site. Pursuing a divisional in Australia in 2026, seven years after the PCT, indicates the family is being actively maintained rather than allowed to lapse, which is usually a sign that a development or licensing conversation is still live.

Related Concepts

• Angiogenesis – the vessel forming process whose dysfunction these antibodies aim to correct.
• TEK tyrosine kinase – the gene encoding Tie2, the receptor targeted by the claimed antibodies.
• Angiopoietin – the ligand family whose balance at Tie2 determines vessel stability.
• Monoclonal antibody – the therapeutic format claimed, here used as an agonist rather than a blocker.
• Faricimab – the approved bispecific that validated this pathway commercially.
• Institute for Basic Science – the co-applicant and host of the Center for Vascular Research behind the work.

---

AU 2026202090 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.