The invisible battle behind RNA delivery

If you've had an mRNA vaccine, you've already experienced one of the most impressive accomplishments in modern medicine: scientists packaging fragile genetic instructions inside tiny lipid bubbles, called lipid nanoparticles (LNPs), to deliver them safely into your cells. But while these microscopic capsules have transformed how we fight viruses, they still face a frustrating problem: most of the RNA they carry never reaches its destination inside the cell.

A new study published in Nature Communications by Johanna M. Johansson, Hampus Du Rietz, Hampus Hedlund, Hanna Eriksson, and Anders Wittrup at Lund University in collaboration with AstraZeneca's BioPharmaceuticals R&D teams in Sweden and the U.S., dives deep into this mystery. Using a combination of live-cell and super-resolution microscopy, the team watched, in real time, how RNA-loaded nanoparticles behave once they enter cells, and uncovered a series of invisible roadblocks that explain why these vehicles are so inefficient at delivering their genetic cargo.

Lipid nanoparticles have been hailed as molecular couriers. They carry snippets of genetic code, either mRNA (used in vaccines and gene therapy) or siRNA (used to silence disease-causing genes), wrapped in a fatty shell. Once injected into the bloodstream, they travel to target cells, fuse with their outer membrane, and ideally, release the RNA into the cell's interior (the cytosol), where it can do its job.

But “ideally” is doing a lot of heavy lifting. While LNPs work well in the liver, where they were first approved for treating rare genetic diseases, they struggle in most other tissues. The RNA gets trapped inside bubble-like compartments called endosomes, internal mailrooms where the cell sorts and digests incoming material. For the RNA to work, it needs to escape these compartments into the cytosol, something that rarely happens.

Until now, scientists knew this escape step was the main bottleneck. What they didn't know was why it's so inefficient. That's where Johansson and colleagues stepped in.

Using state-of-the-art microscopes, the Lund team labeled both the RNA and the lipid shell of LNPs with fluorescent dyes. This allowed them to literally watch the nanoparticles move through cells, merge with endosomes, and sometimes, but not often, burst free.

One of their key tools was a molecule called galectin-9, a kind of “damage detector” that glows when an endosome's membrane is breached. When LNPs disrupted an endosome, galectin-9 would light up, revealing a moment of membrane damage, potentially a portal for RNA escape.

But the researchers quickly realized something surprising: not all damaged endosomes actually released RNA. In fact, many ruptured endosomes contained no detectable RNA at all. It seemed that even when the nanoparticles broke through cellular barriers, their genetic cargo often failed to come along for the ride.

By tracking both components of the LNP, the ionizable lipid (which helps the particle fuse with membranes) and the RNA payload, the team discovered that these two ingredients often go their separate ways once inside the cell. As the endosome matures, the nanoparticle starts to fall apart, and the lipids and RNA drift into different sub-compartments.

In some cases, the lipid accumulates at the endosomal membrane, causing damage or “leaks,” while the RNA remains trapped elsewhere. It's as if the delivery truck crashes through the warehouse wall, but the package stays inside the box.

This explains why the researchers saw so many “damaged” endosomes that were strangely empty, the lipid could still harm the membrane, even after its RNA payload had wandered off.

Cells don't take membrane damage lightly. They have an inbuilt repair system, known as the ESCRT machinery, that rushes in to patch holes and seal leaks before too much escapes. Johansson's team found that this repair system kicks in quickly when LNPs disturb the endosomal membrane, sometimes so quickly that it stops the RNA from escaping altogether.

When the researchers silenced key ESCRT genes, more damage signals appeared in the cells, suggesting that the repair machinery was indeed limiting RNA delivery. The cell, in essence, was fighting back against the therapeutic invasion.

Even when everything went right, the LNP entered the cell, reached the right compartment, caused a rupture, and avoided rapid repair, only a small fraction of the RNA actually escaped into the cytosol. The process was fast, often happening within seconds, but incomplete. Many nanoparticles remained tethered to the endosomal membrane, leaking out only a portion of their payload before being sealed off or digested.

What's more, larger mRNA molecules fared worse than smaller siRNA ones. While up to half of siRNA molecules could escape from damaged endosomes, fewer than one in five mRNA-containing endosomes released detectable cargo. The bigger, bulkier strands simply had a harder time squeezing through the same tiny holes.

The team's microscopic images are almost cinematic. They show glowing RNA specks clustering near membranes, flickering as the endosome wall ruptures, and then fading as some RNA drifts into the cell's interior. In other frames, bright lipid patches appear like fireflies on the membrane's edge, areas where the ionizable lipid has accumulated and caused stress.

At times, the lipid and RNA appear side by side, almost touching; in others, they drift apart completely. These visual cues reveal a chaotic dance between particle disintegration, membrane damage, and cellular repair, a dance that determines whether the therapy works or fails.

LNPs are the workhorses of today's RNA therapeutics. They made mRNA vaccines possible, but scientists want to go much further, using LNPs to deliver RNA that repairs genes, kills cancer cells, or regenerates damaged tissues. To reach those goals, LNPs must reliably deliver their cargo into many different cell types, not just liver cells.

By pinpointing exactly where and why RNA delivery fails, Johansson and colleagues have provided a roadmap for improvement. Their findings suggest that to make LNPs more effective, scientists need to prevent lipid–RNA separation inside endosomes, so both parts stay together until release, control the timing and extent of membrane damage, balancing escape with cell safety, and temporarily modulate the cell's repair system, giving RNA a better chance to get out before the ESCRT machinery patches the hole.

In other words, future generations of LNPs will need to be smarter couriers, able to sense when to open the package and where to deliver it, without triggering the cell's emergency response.

The approach of the team, which combines live-cell imaging with biophysical analysis, could soon be used to test new types of nanoparticles, including those made with different lipids like ALC-0315 (used in Pfizer's COVID-19 vaccine) or SM-102 (used by Moderna). These variants were engineered to improve RNA delivery, but no one has yet seen exactly how they perform inside living cells at this level of detail.

Ultimately, this line of research could usher in a new era of rational design for RNA delivery systems, moving away from trial-and-error chemistry toward evidence-based engineering guided by what actually happens inside the cell. What Johansson and her colleagues have shown is that inside every cell, a microscopic battle plays out between the nanoparticle, its fragile RNA cargo, and the cell's own defenses. Most of the time, the cell wins. But thanks to this research, we now know what needs to change for the RNA to stand a fighting chance.

If you want to learn more, read the original article titled "Cellular and biophysical barriers to lipid nanoparticle mediated delivery of RNA to the cytosol" on Nature Communications at http://dx.doi.org/10.1038/s41467-025-60959-z.