Size-dependent cytotoxicity of silver nanoparticles in human lung cellsthe role of cellular uptake, agglomeration and Ag release

The 10 nanometer silver nanoparticles are more toxic than the larger ones. The data confirm this. But here is what makes it interesting: when Gliga and colleagues measured how much silver actually got inside the cells, the numbers were nearly identical across sizes. The coating didn't change it either. Whether citrate or PVP, it didn't matter. So if uptake isn't the explanation, what is? Silver nanoparticles are embedded in a wide range of products — medical devices, cosmetics, clothing, room sprays, and even food packaging — with roughly 320 tons manufactured each year. Their antimicrobial properties drive that demand. However, the lungs are a primary exposure route, and the toxicology literature is complex. Reviews identify oxidative stress, apoptosis, and genotoxicity as common outcomes, but individual studies disagree sharply. One group found reduced cell viability at 10 micrograms per milliliter after just one hour. Another reported no toxicity at all up to 100 micrograms per milliliter. Kittler and colleagues found toxicity increased with storage time, correlating with silver ion release. Liu and colleagues found smaller particles to be more toxic, while Kim and colleagues found the opposite. The inconsistency isn't random; it reflects the fact that particles vary in synthesis, size distribution, and surface chemistry. The cell lines differ as well, and critically, almost no one was characterizing what the particles actually became once they entered cell medium. That is the problem Gliga and colleagues set out to fix. They assembled five particle types: citrate-coated silver nanoparticles at 10, 40, and 75 nanometers; a 10 nanometer PVP-coated particle; and a 50 nanometer uncoated powder. All sourced commercially and characterized by transmission electron microscopy in water to confirm primary particle size. Then — and this is what separates this study — they measured agglomeration directly in the actual cell medium, bronchial epithelial cell growth medium, at 0, 4, and 24 hours using photon cross-correlation spectroscopy. They also tracked sedimentation with ultraviolet-visible spectroscopy. What they found is that all particles agglomerate in cell medium, but the patterns depend on coating. The 10 nanometer citrate particles showed a trimodal size distribution — peaks at roughly 10, 100, and 1000 nanometers — and those larger modes grew and then sedimented by 24 hours. The 10 nanometer PVP particles remained largely below 10 nanometers and stayed stable. Coating clearly shaped stability. But as Gliga and colleagues note, that stability difference did not translate into a difference in what happened to the cells. That discovery is yet to come. The cytotoxicity result is stark. After 24 hours, only the 10 nanometer particles — both citrate and PVP coated — reduced cell viability. The Alamar Blue metabolic assay showed significant reductions for both 10 nanometer formulations at the highest doses. The lactate dehydrogenase assay, which measures membrane damage, confirmed it: significant effects for the 10 nanometer citrate particles, and highly significant for the 10 nanometer PVP particles at 50 micrograms per milliliter, with a p-value below 0.0001 for the latter. The 40, 75, and 50 nanometer uncoated particles did nothing in either assay. At four hours, nothing was toxic. By 24, the size split was unambiguous. So the team went looking for the reason. The first candidate was uptake — maybe more 10 nanometer particles were getting inside the cells. Transmission electron microscopy showed all particle types being taken up into membrane-bound endo-lysosomal vesicles. There was no nuclear localization for any particle. No clear difference in compartmentalization between sizes. Quantification by atomic absorption spectroscopy, which measures silver directly, returned cellular doses in the range of about 2 to 10 picograms per cell after 4 hours. The highest uptake was for the uncoated 50 nanometer particles. The 10 nanometer citrate and the 75 nanometer citrate particles had similar cellular silver loads, around 3 picograms per cell each. Uptake inhibitor experiments at 4 degrees Celsius dropped uptake to a few percent for both 10 nanometer and 75 nanometer particles in parallel. There was no unique entry route for the toxic particles. Uptake was not the answer. The second candidate was oxidative stress. Reactive oxygen species — chemically reactive molecules that can damage DNA and proteins — were measured with the DCFH-DA assay after 24 hours. None of the silver nanoparticles produced a significant increase. The positive control, tert-butyl hydroperoxide, produced a 2.8-fold increase over unexposed cells. The nanoparticles produced nothing detectable. The third candidate was DNA double-strand breaks. The gamma-H2AX foci assay is specific for double-strand break damage — it lights up when chromosomes are seriously fractured. It was negative for all silver nanoparticles at both 4 and 24 hours. Etoposide, the positive control, produced foci. The nanoparticles did not. But here the story gets genuinely strange. The alkaline comet assay — a different DNA damage test, sensitive to single-strand breaks and alkali-labile sites — showed that all silver nanoparticles, regardless of size or coating, produced significant DNA damage after 24 hours at 10 micrograms per milliliter. All of them. The cytotoxicity was size-specific. The DNA damage was not. This means the mechanisms are separate. Some form of genotoxicity is happening across the board, but only the smallest particles are killing cells — and the usual explanations have all been ruled out. The answer the paper proposes comes from measuring silver dissolution. Using atomic absorption spectroscopy on supernatants from particle dispersions incubated in cell medium, Gliga and colleagues found that 10 nanometer particles released approximately 24 weight percent of their total silver as dissolved species after 24 hours. The larger particles released only 4 to 7 weight percent over the same period. This represents nearly a six-fold difference in fractional release. The physical reason is surface area. For the same mass of silver, 10 nanometer particles present a one hundred ninety-six-fold greater number of particles — and proportionally greater surface area — compared to 40 nanometer particles, and a one thousand one hundred forty-fold greater number compared to 75 nanometer particles. More surface exposed to the medium means more dissolution. This is not a subtle effect. It is a geometric inevitability that scales non-linearly as particles shrink. Now add the Trojan horse logic. The particles enter cells intact, as transmission electron microscopy confirmed, and they end up inside acidic lysosomal compartments. In that environment, dissolution continues. The 10 nanometer particles, with their enormous surface-to-volume ratio, release far more silver ions locally than the larger particles do. Those ions — at high local concentration inside the cell — can disrupt biological function in ways that extracellular ion exposure cannot replicate. Gliga and colleagues tested this directly: they collected the supernatant from 50 micrograms per milliliter dispersions of 10 nanometer particles after 24 hours — the released fraction, containing whatever dissolved silver was already free in medium — and exposed cells to it. There was no cytotoxicity. The ions floating around outside the cell, at the concentrations generated by even the most dissolving particles, were not enough. It is the intracellular release that counts. The lysosomal fluid experiments using artificial lysosomal fluid at pH 4.5 showed very little measurable release — less than 2 percent — but the authors attribute this to agglomeration, chloride complexation, and sedimentation during sample processing rather than a genuine absence of dissolution. The intracellular case is more complex than a simple pH experiment, but the broader argument holds: particles deliver silver into the cell, and then release toxic ionic species in a confined intracellular space where concentrations can build. What Gliga and colleagues demonstrate, finally, is that particle characterization in stock solutions is not sufficient for nanotoxicology. The same particles behave differently in biological media. They agglomerate on timescales of hours, they sediment, and they dissolve — and all of those processes are time-dependent and coating-dependent and size-dependent in ways that only emerge when you measure them in the relevant environment. Without that, inter-laboratory toxicity data will keep disagreeing because the particles being compared are not actually the same once they enter the experiment. For silver nanoparticles specifically, the clearest takeaway is this: it may not be the particle that is toxic at all. It is the ion it releases once it is already inside. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.