The tiny fat bubbles that carry mRNA vaccines into our cells work remarkably well — and they also tend to spark inflammation. That side effect is manageable for a one-or-two-dose shot, but it has blocked many other uses of the same delivery trick, especially for people who already live with inflammation from cancer, diabetes, or lung injury.
A UC Berkeley–led team now reports a new kind of lipid nanoparticle — a microscopic fat droplet that ferries genetic medicine — built so its electrical charge flips with acidity. In blood, which is near neutral, the particles stay less inflammatory. In the acidic pockets inside cells, they go positive so they can still unload messenger RNA (mRNA), the temporary instruction slip cells use to make a protein. The work is preclinical mouse and human-blood-cell research published in Nature Nanotechnology, not a new approved drug.
Why today’s fat bubbles pick a fight
Lipid nanoparticles (LNPs) became household terms during COVID-19. They wrap fragile mRNA in a fatty shell so it survives the bloodstream and reaches cells. The engine of that shell is usually an ionizable lipid — a fat molecule whose charge can change with pH. At the low pH used to load cargo, and again inside acidic endosomes (the cell’s internal sorting pouches), that lipid goes positive. Positive charge helps the particle grab nucleic acids and break out of the endosome so the mRNA can be read.
The catch is that the same positive character also activates inflammatory alarm systems. Studies have linked conventional LNPs to pathways such as TLR4 (a sensor that often fires during bacterial infection), the complement cascade (a blood-based immune amplifier), galectin-8 (a protein that marks damaged endosomes), and platelet-activating factor (PAF), a signaling molecule tied to inflammation. For a vaccine, a short burst of immune noise can be acceptable. For repeated doses, high doses, or patients who already have inflamed tissues, it can be dangerous.
Niren Murthy, a UC Berkeley bioengineering professor and Innovative Genomics Institute member, has described the bind in simple terms: for about thirty years, designers had to trade toxicity against delivery efficiency. That trade-off slowed work beyond vaccines.
A molecular switch on the fat itself
The new paper’s answer is a charge-switching ionizable lipid, or S-lipid. Chemically, each S-lipid carries both a carboxylic acid and an amine in its head group. That combo lets the molecule switch charged states between roughly pH 7.4 (near-neutral blood, where the particle behaves as negatively charged or less inflammatory) and about pH 4.0 (acidic conditions needed for packaging cargo and for endosomal release).
Dengpan Liang, a postdoctoral researcher in Murthy’s lab and a co-first author, put the everyday picture this way: the switchable nanoparticles stay positive when acidity demands it — while loading mRNA and escaping the endosome — then flip toward a negative state in circulating blood, which helps blunt toxicity without abandoning delivery power.
The team screened a large library (on the order of 144 designs) and refined top hits. Good performers needed a several-carbon linker between the amine and carboxyl groups, branched hydrophobic tails, and other design rules that made some lipids thousands of times better than weak cousins in cell tests. The particles they call switchable nanoparticles (SNPs) still encapsulate mRNA and plasmid DNA efficiently and can match traditional LNPs on delivery in several models — while skipping the inflammatory signature those older particles usually leave.
What the lab tests actually showed
In mice, several SNPs delivered luciferase mRNA (a glow-in-the-dark reporter used to measure delivery) as well as or better than a classic benchmark LNP known as MC3-DLin. Many of the first-generation particles headed mainly to the liver; others showed more spleen preference. With human erythropoietin (EPO) mRNA — a protein-replacement style cargo — top SNPs raised blood EPO to levels comparable to the traditional control.
The inflammation story is sharper. When mice were first challenged with LPS (a bacterial molecule that creates pre-existing inflammation), traditional LNPs spiked cytokines such as IL-6 and IL-1β. An SNP called E20 did not worsen that inflammatory state in the same setup. In a six-LNP-style panel on human peripheral blood mononuclear cells (immune cells from blood), the switchable particles were the only class that did not trigger the cytokine burst the other formulations produced — while still delivering luciferase mRNA efficiently.
Pathway assays lined up with the charge story. SNPs did not activate TLR4 or complement the way several traditional LNPs did. They also showed little galectin-8 recruitment and did not drive PAF release from human blood cells in the reported tests. Mechanistically, the researchers suggest the particles can exit productive endosomal compartments without ripping membranes the way harsher cationic formulas do.
DNA delivery highlighted the safety gap. Mice given plasmid DNA in a traditional LNP died within a day in the reported experiment; mice given the same cargo in E20-SNPs survived a week of observation and still expressed the reporter gene. That is a mouse result under specific doses — not a green light for human DNA therapies — but it shows how much inflammation can dominate outcomes.
A harder test: treating inflamed lungs
Acute lung injury is a brutal setting for traditional LNPs because the tissue is already inflamed. The team loaded SNPs with IL-22 mRNA, which codes for a protein known to help protect lung tissue in LPS injury models, and compared them with several traditional LNPs after an LPS challenge in mice.
SNP-treated animals showed better rescue of lung leakiness and cell infiltration than animals treated with the conventional particles, even though IL-22 delivery efficiency was similar. In other words, the advantage looked less like “more drug” and more like “the vehicle did not pour gasoline on the fire.” That is exactly the clinical intuition Murthy and colleagues hope will matter for people who need nucleic-acid drugs on top of existing inflammation.
What this is — and is not — ready for
Out of the gate, the group points to liver protein replacement as a near-term research target. One example they discuss is ornithine transcarbamylase (OTC) deficiency, a rare inherited disorder in which the body cannot clear ammonia well. Murthy has said that if switchable LNPs could deliver the right mRNA with low toxicity on a chronic schedule, diseases like OTC deficiency could become treatable in a new way. That remains a research aspiration: there is no clinical trial result in this paper showing a human OTC cure.
Limits are explicit. First-generation SNPs mainly go to liver and spleen. The team wants particles that reach more tissues and lipids that break down faster after they deliver their cargo, so material does not linger. Murthy noted they have made on the order of twenty strong lipids so far and expect the design rules to yield many more. Chronic dosing, long-term accumulation, and human safety still have to be proven outside these preclinical models.
The study was co-led by Murthy (Berkeley), Aijun Wang (UC Davis / Shriners Children’s Northern California), and Hesong Han (Berkeley), with Dengpan Liang among the co-first authors. Collaborators spanned Stanford, Lawrence Berkeley National Laboratory, Chongqing University, and the Chan Zuckerberg Biohub. Berkeley Engineering highlighted the work in early October 2026; the peer-reviewed paper is in Nature Nanotechnology (“Charge-switching ionizable lipids lower the toxicity of lipid nanoparticles”).
For readers, the everyday takeaway is simple. mRNA vaccines proved that fat nanoparticles can teach cells to make useful proteins. Inflammation from those particles has been the brake on many other medicines. Charge-switching S-lipids are an attempt to keep the delivery power while dialing down the fight — a clever chemistry fix for a thirty-year trade-off, still early, but aimed at a problem patients and drug developers both feel.
As an Amazon Associate I earn from qualifying purchases. If you buy through links on this page, I may earn a commission at no extra cost to you.
Further reading
Breaking Through: My Life in Science — Katalin Karikó’s memoir of the decades-long mRNA research that made modern vaccines possible — useful context for why delivery chemistry still matters.
The Messenger: Moderna, the Vaccine, and the Business Gamble That Changed the World — Wall Street Journal reporter Peter Loftus on how mRNA lipid-nanoparticle vaccines moved from lab bet to pandemic-scale reality.