Two Jobs for One Viral Coat Particles That Find a Single Immune Cell Type and Switch It On

This application covers virus-like particles and retroviral vectors that carry two different engineered spikes at once: one that steers the particle to a chosen cell surface marker, and one that delivers a signal, typically a cytokine, to the cell it lands on. Both are built on envelope glycoproteins taken from the Paramyxoviridae family. It was filed by four French public research bodies whose joint infectious disease centre in Lyon has been engineering targeted viral vectors for two decades.

The Problem

Immune cells are separated into functionally distinct subtypes by the proteins they display on their surface, and much of therapy design amounts to wanting to turn one subtype up or down without touching the others. The specification states the gap bluntly: there is so far no technology available that allows the selective activation or deactivation of only a distinct cell subtype, especially in a living body.

Cytokines are the obvious tool and the obvious disappointment. Given systemically they reach every cell carrying the matching receptor, which is why unresolved toxicity has kept cytokine treatments in a supporting role. Fusing a cytokine to a tumour targeting antibody, or loading it onto nanoparticles, improves the aim but introduces new faults: the constructs are less stable, they are cleared quickly, release from particles depends on local pH, and once delivered the cytokine still stimulates effector and suppressive immune cells alike.

Gene delivery has a mirror image of the same problem. Resting T cells, B cells and haematopoietic stem cells resist transduction, and the usual vector coat, from vesicular stomatitis virus, only works once T cells have been activated through their antigen receptors, which changes what those cells become. Vectors displaying interleukin 7 alongside that coat can transduce resting T cells, but hit CD4 and CD8 cells indiscriminately. Published CD4 targeted vectors reached freshly isolated resting T cells at under ten per cent efficiency even at high particle doses.

What This Invention Does

The inventors’ finding is that separating the two jobs, and putting both on the same particle, works far better than either alone. Claim 1 defines a pseudotyped retrovirus-like particle or retroviral vector carrying three things. First, a cell targeting fusion protein: a protein derived from the G or H envelope glycoprotein of a paramyxovirus, joined to a targeting domain. Second, a modulating fusion protein: the same glycoprotein backbone, joined instead to a functional domain. Third, a glycoprotein derived from the F envelope protein, which is what actually fuses the particle to the cell membrane. In the claim as filed, the G or H derived protein lacks part of its cytoplasmic region.

Those truncations matter. In nature G or H handles attachment and F handles fusion, and both carry cytoplasmic tails sized for their own virus. Shortening the tail, by 34 amino acids on a Nipah G protein and 15 on a measles H protein in the worked examples, lets the spikes be picked up by a retroviral core. A second set of mutations blinds the glycoprotein to its natural receptors, so the only address the particle recognises is the one the engineers bolted on.

The targeting domain can be a DARPin, a single chain antibody fragment or a peptide, aimed at markers including CD3, CD4, CD8, CD19, CD20, CD56, CCR5 or CXCR4. The preferred pairings are a DARPin specific for CD4 and a single chain fragment specific for CD8. The functional domain is preferably a receptor ligand, most often a cytokine and most often interleukin 7, though the list runs through other interleukins, interferons, tumour necrosis factor, growth factors, hormones, chemokines and apoptosis ligands such as Fas ligand. The reported result is that particles displaying both a T cell targeting domain and a stimulatory cytokine selectively activate the targeted subset, in mixed cultures and in a mouse carrying a human blood system, and deliver a packaged gene into those same cells.

Key Features

• Two spikes, two jobs. Targeting and signalling are carried by separate fusion proteins on the same particle, rather than compressed into one molecule.
• Glycoproteins blinded to their own receptors. Mutations stop the paramyxovirus G or H protein binding its natural receptor, so the attached targeting domain decides where the particle goes.
• Shortened cytoplasmic tails. Removing part of the tail of the G, H and F proteins allows paramyxovirus spikes to be incorporated into a retroviral particle.
• A choice of viral chassis. The glycoproteins may come from the Morbillivirus genus, including measles virus, or the Henipavirus genus, including Nipah virus, Cedar virus and Hendra virus.
• Signal with or without a payload. The same architecture works as a particle that only modulates the target cell, or as a vector that also delivers a gene into it.
• An open functional menu. The modulating domain is claimed broadly as a receptor ligand, covering stimulation, suppression, differentiation or induced cell death depending on what is fused.

Who Is Behind It

The four applicants are Centre National de la Recherche Scientifique, France’s national research agency, Ecole Normale Superieure de Lyon, the Institut National de la Sante et de la Recherche Medicale, the national biomedical research institute known as INSERM, and Universite Claude Bernard Lyon 1. Those four bodies jointly run the International Centre for Infectiology Research in Lyon, where the vectorology group behind this work sits. Joint ownership across several public institutions is standard for French academic patents and reflects the mixed staffing of the laboratory.

Six inventors are named: Caroline Costa Fejoz, Els Verhoeyen, Francois-Loic Cosset, Ruben Bender, Christian Buchholz and Qi Zhou. Cosset and Verhoeyen lead the Lyon vector engineering effort; Buchholz heads molecular biotechnology and gene therapy research at the Paul-Ehrlich-Institut in Germany, and his group has published extensively on receptor targeted lentiviral vectors, including the CD4 and CD8 targeted vectors this application builds on.

The priority chain is long. This is a divisional of Australian application 2023258359, itself a divisional of 2020203813, itself a divisional of 2017253364, which was the Australian national phase of PCT/EP2017/059435 filed on 20 April 2017 and published as WO 2017/182585. That application claims priority from European application EP 16305468.7, filed on 21 April 2016.

Why It Matters

The field has moved quickly since that priority date. Generating chimeric antigen receptor T cells inside the patient, rather than harvesting cells, engineering them in a facility and infusing them back, is now one of the most actively pursued goals in cell therapy, because it would turn a bespoke manufactured product into something closer to an injection. That approach depends entirely on vectors that can find one lymphocyte subset in whole blood and enter it while it is resting, which is exactly the pair of problems this specification sets out to solve at once.

The Nipah based versions carry an advantage the specification calls out: because human populations have very little pre-existing antibody against Nipah virus, particles coated in its glycoproteins are less likely to be neutralised before they arrive. Measles is the better characterised chassis but runs into widespread vaccine derived immunity, so having both genera inside the claims hedges the risk.

The filing history is the other tell. Three successive divisionals in Australia, alongside a family pursued through Europe, the United States, Brazil, Canada, China, Korea and Singapore, is what an institutional portfolio looks like when the platform is treated as licensable rather than as a one-off result. A decade of maintenance costs on a family this size is a bet that someone will want a licence to put a cytokine and a postcode on the same particle.

Related Concepts

• Chimeric antigen receptor – the therapeutic payload these targeted vectors are being developed to deliver in the body.
• Pseudotyping – giving one virus another virus’s surface proteins, which the whole design rests on.
• Cytokine – the class of signalling molecule most often used as the functional domain.
• Single-chain variable fragment – the antibody derived targeting format used for the CD8 specific version.
• Viral vector – the broader gene delivery category the particles belong to.
• Gene therapy – one of the three uses, with immune therapy and vaccination, named in the claims.

---

AU 2026202115 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.