Next-generation vaccines that look like viruses – but are not

They were fast, effective and helped to end the pandemic. But even the revolutionary COVID-19 vaccines had their limitations.

The secret behind them was a molecule called mRNA – a kind of genetic instruction that tells our cells which proteins to make. mRNA vaccines deliver the recipe for a protein from a pathogenic organism, training the immune system to recognise and attack the pathogen – without ever encountering the real thing.

“The first generation of mRNA vaccines was a breakthrough,” explains Adam Sander Bertelsen, Professor, Department of Immunology and Microbiology, University of Copenhagen, Denmark. “They were developed rapidly, were highly effective at preventing severe disease and played a critical role in ending the COVID-19 pandemic. But they were not perfect. The immune response faded more rapidly than hoped. And the immune system did not always recognise the soluble proteins they encoded.”

Now, Adam Sander Bertelsen and colleagues think they have found a way to unlock the full potential of mRNA – by teaching cells to build virus-like particles (VLPs). The result could be a stronger, longer-lasting immune response – and a modular platform adaptable to virtually any infectious disease.

To test the idea, the team used a notoriously difficult target: malaria.

“The goal was not to solve malaria – it was to prove that the concept worked,” Adam Sander Bertelsen explains.”

A candidate targeting Nipah virus – a known pandemic threat – is already in preclinical development.

From setback to big idea

Before Adam Sander Bertelsen imagined mRNA vaccines that look like viruses, he had already spent years working on their protein-based cousins – VLPs. These engineered shells mimic the shape of real viruses and can provoke strong immune responses without any risk of infection.

When the COVID-19 pandemic struck, Adam Sander Bertelsen and colleagues helped to develop a Danish protein-based VLP vaccine in collaboration with Bavarian Nordic. The vaccine performed well in Phase III trials, showing non-inferiority to commercial mRNA vaccines – even without using an adjuvant.

“We got further than we ever expected,” says Adam Sander Bertelsen.

But the landscape had shifted: mRNA vaccines could be updated faster. Regulatory authorities now expected annual reformulations and large-scale commercial production – requirements that were beyond the reach of an academic laboratory.

“We were on the verge of bringing it to market as a viable product and believed it had strong potential to become commercially competitive. However, the sudden emergence of new requirements for annual updates to COVID-19 vaccines made moving forward with commercialisation extremely challenging for Bavarian Nordic. So this was a massive downer.”

The setback triggered reflection – and ultimately, a conceptual breakthrough.

“We realised that the platform worked – but that we would need to take it much further.”

Adam Sander Bertelsen began to envision a hybrid approach: using mRNA to encode not just the antigen but the entire VLP structure. If this worked, the body itself could assemble the vaccine from scratch – combining the speed and adaptability of mRNA with the durability and structure of VLPs.

“I thought: what if you had the flexibility and speed of mRNA delivery but with VLP technology inside – so you get the strongest possible and maybe even a long-lasting response?”

This new idea became the foundation of the next-generation vaccine platform.

How the body builds vaccines

To many, the idea of combining mRNA delivery with VLP design seemed far-fetched. Adam Sander Bertelsen’s peers were sceptical. How could a single strand of mRNA instruct cells to build not just a viral protein but an entire nanoparticle that mimics the shape of a virus?

“Some people said, ‘Adam, that sounds a bit strange’. It is not like everyone said, ‘Great, let’s go for it.’”

But Adam Sander Bertelsen could not let it go. His years of experience with VLPs told him that structure matters – and if the body could assemble that structure itself, it could transform vaccine design.

“These VLPs work well, but they are complicated to manufacture. What if mRNA could encode both the antigen and the virus-like structure – so the body does the building? We thought: we will just make it. Then we will see whether it works.”

Early experiments used a system called Tag/Catcher – molecular “Velcro” that links the antigen to a structural protein. One mRNA strand encoded both parts. The idea: inject the recipe and let the body build the VLP itself.

“This is like giving the body both the manual and the materials – and then it handles the rest. It is faster, cheaper and far more effective – if it works.”

The initial attempt did not. The VLP scaffold they chose failed to form particles effectively. But instead of giving up, the team screened a library of particle-forming proteins, testing which ones could reliably self-assemble into a VLP with the antigen. Eventually, they found several that worked.

That is when the idea became real. For the first time, mRNA was not simply instructing the body to make a loose protein – it was building a shaped particle, one that the immune system interprets as a virus.

“The real breakthrough came when I thought: ‘I cannot just ignore mRNA technology.’ It worked. Maybe not quite as well as I had hoped – but it worked. And imagine if you could combine the two approaches? Maybe that would solve the biggest problem with mRNA – that the immune response does not last.”

Malaria: a brutal test case

With a working VLP-mRNA system in hand, the team faced a new question: would it actually generate strong immunity? To find out, they needed a target. And not an easy one.

“I started my career in malaria vaccine research, so that was a natural place to begin,” says Adam Sander Bertelsen. “But more importantly, malaria is really difficult. The goal was not to solve it – but to prove that the concept worked. If it worked there, it could work anywhere.”

Malaria was not the end goal – but it was a brutal test case. The antigen they chose was Pfs25, a protein known to be especially challenging for vaccine development.

The results were striking. The mice produced VLPs using mRNA alone, and the immune response was not only strong but lasted longer than that of conventional mRNA vaccines.

“The VLP-mRNA vaccine triggered stronger immune responses at lower doses – enabling more mice to be protected with the same production volume,” explains another lead author, Cyrielle Fougeroux, Postdoctoral Fellow, Department of Immunology and Microbiology, University of Copenhagen, Denmark.

This is especially important in global health. That dose-sparing effect has profound implications for vaccine access and affordability.

The secret, Adam Sander Bertelsen believes, lies in how the immune system sees the threat. Traditional mRNA vaccines generate soluble antigens – protein blobs that float around the body. But the immune system is not good at spotting blobs.

“We did not evolve to recognise blobs. We evolved to recognise shapes that look like viruses – clusters of proteins arranged in patterns, like the shell of a real virus.”

The VLP-mRNA vaccine gave the body exactly that: a virus-shaped package built from within.

“That viral shape triggers a stronger and longer-lasting immune reaction, and so it reacts more strongly and for longer.”

Stronger immunity, less vaccine needed

Once the malaria test succeeded, the team pushed further. Laboratory experiments confirmed that dendritic cells – key players in triggering both antibody and T-cell responses – took up the VLPs more efficiently.

In mice, the result was a stronger, longer-lasting immune response than conventional mRNA vaccines – and it worked at significantly lower doses.

“It is like getting more bang for your buck,” says Cyrielle Fougeroux.

Lower doses could transform vaccine equity in low-resource settings. The platform showed another advantage: thermal stability. That makes it more viable in regions with limited temperature-controlled supply chain (cold-chain) infrastructure.

At this stage, the platform was not just proving effective. It was becoming modular and scalable.

“The clever part is that you can use one particle and just change the antigen,” adds Cyrielle Fougeroux. “It is like swapping the head on a Lego figure – the body still recognises a virus but with a new face.”

The researchers began testing the platform with antigens from respiratory syncytial virus, influenza virus and SARS-CoV-2. Because the structural (capsid) proteins stay the same, only the antigen needs to change – a plug-and-play design that could dramatically speed up vaccine development.

“A single day’s experiment can reveal which antigens work with the system. In a pandemic, this speed can save months – and potentially thousands of lives.”

One platform, many viral targets

With the malaria results in hand, Adam Sander Bertelsen’s team looked ahead. The next target: Nipah virus, a deadly zoonotic disease and top WHO priority.

Their approach? Develop two versions of the vaccine: one using the new mRNA-VLP technology and the other based on protein-manufactured VLPs – both using the same antigen display.

“There is a real possibility that we could launch a first-in-humans trial based on the Nipah vaccine candidate,” says Adam Sander Bertelsen. “Both platforms use the same VLP display – just delivered in different ways.”

So why develop both? Both platforms mimic viruses to spark immunity – but they have different strengths. mRNA versions often trigger stronger T-cell responses, which could be vital for diseases such as Nipah virus infection. Protein vaccines offer more control in the laboratory and can be fine-tuned for stability.

“One key finding is that mRNA delivery of the same antigen and particle produces a much stronger cellular response,” explains Adam Sander Bertelsen. “So if the disease demands T-cell activation, that might tilt the decision.”

And then there is the question of logistics.

“In some parts of the world, you might actually prefer a protein vaccine,” notes Adam Sander Bertelsen. “Because it is easier to distribute without cold chains.”

Ultimately, the goal is not to choose one winner – but to build a versatile toolbox.

€8 million boost for Nipah trials

The Nipah project did not just mark a scientific milestone – it became the launching point for a major European investment. In 2024, Adam Sander Bertelsen’s team secured a €8 million EU Horizon Europe grant to accelerate the development of the vaccine platform, including both protein-based and mRNA-based versions.

Preclinical studies are now underway, and plans are in motion for first-in-humans trials, depending on the results.

“We cannot say we are doing anything until we have shown this in humans,” says Adam Sander Bertelsen. “Mice are great for generating ideas, but their immune systems are not quite like ours.”

If it succeeds, the implications are sweeping: stronger protection, faster development, lower costs – and no need to start from scratch for each new threat.

“If the technology is tested and approved for humans, development will go much faster the next time.”

But Adam Sander Bertelsen has not forgotten the near miss during COVID-19. Being just a few months too late to join the vaccine race left a mark – but also clarified the mission. As he says, the idea held up – and the setback only sharpened the vision.

“That is when we realised that the platform is viable – we just needed to push it further. The setback made us sharper.”

The concept has come full circle. Adam Sander Bertelsen wanted a platform that was fast, flexible and potent – and now, he believes it is ready to be put to the test. And the core idea remains simple: make it look like a virus. Let the body do the rest.