Modified polymer could help people with PEG allergies

Biological drugs such as gene therapies and messenger RNA (mRNA) vaccines frequently rely on lipid nanoparticles (LNPs) coated in a layer of polyethylene glycol (PEG) to deliver medications to the intended target. But, likely from general exposure to PEG through everyday products such as food and cosmetics, a small percentage of people have developed PEG allergies. A new tweak to PEG’s structure could allow it to evade the anti-PEG antibodies (APAs) that cause the allergic reactions (J. Am. Chem. Soc. 2025, DOI: 10.1021/jacs.5c02716).

LNPs are greasy lipid structures, so they are poorly soluble in the water-based environment of the body. Incorporating a surface layer of polar PEG molecules creates a watery shell around the particle, enabling it to move freely through the blood. The PEG layer also protects LNPs from immune proteins, producing a stealth effect that makes the nanocarrier invisible to the immune system, explains Holger Frey, a polymer chemist at Johannes Gutenberg University Mainz. “It's like Harry Potter's invisibility cloak!”

First proposed in the 1970s, this technique of sneaking medications past the immune system is now vital to the success of many drugs, perhaps most notably in mRNA vaccines for COVID-19. But the growing prevalence of APAs is slowly reducing the efficacy of this stealth approach and even triggering allergic reactions in some patients.

According to Frey, most people have some APAs but not enough to cause problems in the body. Around 5% of people experience PEG allergies or faster clearance of the drug from the body, he says, and about 1 in 250,000 people, or 0.0004%, have severe complications like anaphylactic shock. As an increasing number of drugs rely on LNP formulations, developing alternative stealth polymers that won’t trigger reactions in people who experience PEG allergies is useful.

Recent structural analyses suggest these problematic APAs recognize and bind to long regular segments in the PEG chain, around 16 monomers in length. So, rather than starting from scratch with an alternative polymer, Frey used a careful modification of the existing PEG polymer to break up these repeating units and outwit the immune system.

Frey’s team prepared randomized PEG (rPEG) via the same alkoxide-mediated ring-opening reaction that chemists use to make the conventional polymer. But they incorporated into the reaction a second monomer unit—with exactly twice the molecular weight and the same carbon-hydrogen-oxygen ratio as the original monomer—to create a mixture of branched structural isomers. Both of the monomers have the same reactivity, Frey explains. “They are incorporated into the chain at the same rate, which means we have an ideally randomized system.”

As a structural isomer of the original PEG polymer, rPEG should have the same stabilizing and solubilizing effect. But, because rPEG’s polymer chain is interrupted by random branches, the immune system is in theory unlikely to recognize and react to it. The similarity to the existing manufacturing process also means rPEG should prove comparatively straightforward to scale up, adds Frey. His team is already working with specialty chemical manufacturer Evonik to ramp up production.

Nanomedicine specialist Cristina Fornaguera at University Ramon Llull found the team’s approach novel and innovative but believes several questions remain unanswered around the polymer’s long-term behavior in biological systems. “Using a non-conventional PEG is interesting in the sense that the population does not yet have APAs, so this PEG will not be immunogenic in the first instance. However, what will happen with repeated exposure?” she asks. “In addition, no experiments were performed in animals so the conclusion that these polymers will be less allergenic than current available PEGs might be somewhat premature,” she adds.

Frey’s team is already working on a mouse model to validate the immune properties of rPEG. “We believe it is unlikely [that someone would develop an immune response] because the statistics within the chain make it very, very difficult for an antibody to find something it can specifically bind to—there is no regular sequence,” he says.