By Christina Nunez

November 13, 2026 | For decades, scientists have been using the cell line from a crop pest combined with one of its viral foes to produce virus-like particles (VLPs). HighFive cells from BTI derived from the cabbage looper moth, when infected with altered baculovirus, will deliver high yields of proteins that can aid education and research.

As a PhD student 10 years ago, Maren Schubert knew this was an established method for making proteins. But the task in front of her at the time was daunting: She needed to produce 40 different proteins for a high-throughput initial expression screening method. "It takes a lot of time to produce the proteins with the baculovirus system," she says. "One month at least, if you're lucky."

Hoping for faster outcomes, she began experimenting with other options. She found that using plasmids, rather than baculovirus, produced the same yield but was faster, purer, and more flexible. Schubert, who is now a research group leader for VLP-based technologies at the Helmholtz Center for Infection Research in Braunschweig, Germany, has since gone on to refine the method and is presenting her latest research at PEGS Europe in Lisbon, Portugal, November 16–19.

Schubert wasn't the first to experiment with plasmid-based VLP systems, but she has hit on an optimized method that offers advantages over the conventional approach while achieving comparable or even better yields. She likens the plasmid system to a "molecular playground" with LEGO, where components can be assembled to produce the VLP of choice. She and colleagues have used the technique to create harmless replicas of respiratory syncytial virus (RSV), norovirus, Ebola, and other pathogens, exploring how the viruses work and how to defeat them with vaccines and medicines.

The Plasmid Advantage

The baculovirus expression vector system (BEVS) is well established and produces high yields of VLPs. It was used to create GlaxoSmithKline's Cervarix, an early bivalent HPV vaccine still prescribed outside the United States. Why pursue an alternative? The reasons go well beyond saving time. Schubert notes that baculovirus kills its host cells, which interferes with protein quality.

But the most important drawback is the co-expression of baculoviral particles that are difficult to get rid of. The problem is particularly troublesome with VLPs that mimic enveloped viruses such as SARS or hantavirus. Baculovirus will contaminate the VLP, she says, "and you will never get rid of it no matter what purification method you use, because it's inside on the membrane."

With plasmids, by contrast, no viral infection of the host cell is needed. The plasmid encoded with the gene of interest is added via polyethylenimine."The proteins will assemble themselves into these virus-like particles and bud out," Schubert says, drawing a comparison to the human embryonic kidney (HEK) cell system. "It's really straightforward."

Schubert says other labs such as the Fraunhofer Institute for Translational Medicine and Pharmacology ITMP have adopted her plasmid system and are happy with it, but that incentive may be low to change if a lab is already running a working BEVS. "It's sometimes a bit difficult to get people to change a running system for one that is maybe even easier. But in the first place, you have to compare them in your lab to see if they work the same or not," she says. The labs that have tried her system are just setting up an insect cell expression system rather than running one already, she added—the plasmid version is "way simpler" to set up.

One counterintuitive advantage of the plasmid system is that it requires cells at a very high cell passage number: "The higher the cell passage number, the better the cells get," Schubert says. Her lab has cells that have been subcultured over 600 times; as the passage number rises, the HighFive cells get a bit bigger, the doubling time is faster, and they produce more. On the other hand, Schubert acknowledges, the method requires larger amounts of DNA to transfect cells, compared with having a virus that will amplify itself.

Next Goals for Plasmid-Based VLPs

Schubert's group at Helmholtz is involved in two large European Union research projects: one that aims to develop a West Nile vaccine and another seeking a therapeutic antibody that can knock out pox viruses such as Mpox and smallpox. In the latter project, her team is producing smallpox proteins and virus-like particles (VLPs) for antibody analysis and quality control—valuable, since live stocks of smallpox are highly restricted.

Purification remains a hurdle for any complex protein such as VLPs, Schubert notes, and good manufacturing practice (GMP) production is a key hurdle. Plasmid-based systems' flexibility makes them valuable for exploration, but the stability isn't adequate. "If we really want to continue to clinical trials, we have to get to the GMP readiness," she says.

Schubert is also appearing on a PEGS Europe panel about the challenges of experimental data collection for effective and reliable protein design. In theory, artificial intelligence could help predict successful protein designs. But in practice, she says, it's difficult to get comparable data: Sometimes the same method will produce different results across labs for unknown reasons.

Schubert joined a recent effort across more than 70 institutions, detailed in the Journal of Biological Chemistry and dubbed EXPERT (EXpression and Purification Experimental Reporting Template), to define a framework for capturing protein expression and purification data. "It was very interesting to see all the people discussing what kind of data is needed," she says. "I'm curious if we can really manage to collect the data. I'm also curious if we have thought about all the data that we need."