Ozone Emerges as the Leading Driver of Urban Air Toxicity in Barcelona Tests

Researchers have exposed human lung cells directly to the open air of Barcelona, Spain, using a mobile exposure system carried in a van, and the results point to an unexpected culprit. Among all the pollutants measured at a busy harbor site and an urban background location, ozone showed the strongest association with cellular damage, even though particulate matter is conventionally treated as the main driver of urban air toxicity. The study, published in Environmental Science and Pollution Research, combined real-time air quality monitoring with direct cell exposure at two contrasting sites in the city, offering one of the most detailed in situ pictures yet of which components of a real pollution mixture harm lung cells.

Unlike conventional toxicological approaches, which collect particles on filters, extract them into liquids, and then apply those suspensions to cultured cells, the ALI system grows lung epithelial cells on permeable inserts and flows untreated ambient air directly over their upper surfaces. Particles and gases deposit onto the cells just as they would onto the lining of a human lung, and because the cells are exposed to both phases simultaneously, the method can capture the toxicity of reactive gases that are poorly preserved or entirely lost in filter-based sampling.

The mobile setup contained six separate exposure chambers held at 37 degrees Celsius with relative humidity above 90 percent, conditions chosen to keep the cells healthy and promote particle deposition. Four chambers received unfiltered ambient air, while two received air passed through filters that remove more than 99.9999 percent of particles but contain no gas adsorbents, allowing the researchers to separate the effects of the gas phase from those of the particle phase. Cells were exposed for two hours, then incubated for 24 hours before analysis. Human A549 cells, a well-characterized type II alveolar epithelial cell line, served as the biological model, and lactate dehydrogenase (LDH) release was chosen as the primary endpoint because it provides a sensitive quantitative measure of membrane damage and cell death at low levels of injury.

Three campaigns were conducted. Two took place near the Barcelona harbor, downwind of cruise and shipping emissions and deliberately positioned away from road traffic, in late January and early February 2023. A third ran from late September to early October 2022 at an urban background site on the University of Barcelona campus, about 220 meters from the major road Diagonal Avenue. The two sites lie six kilometers apart. During every two-hour exposure, instruments in the same van or at a fixed monitoring station measured ozone, sulfur dioxide, nitrogen oxides, black carbon, particle number concentrations, and the mass of particulate matter in three size fractions, so that the concentrations the cells actually experienced were known precisely.

The toxicity results showed clear damage in many sampling events at both sites. Nine of the 17 sampling events produced LDH values significantly higher than their unexposed controls, and the highest activity occurred during the second harbor campaign. Notably, no significant differences emerged between filtered samples, which exposed cells only to gases, and unfiltered samples, which exposed them to the full mixture of particles and gases. Nor did the particle size cut-off, whether nanoparticles or PM2.5, influence the outcome. Taken together, these patterns suggested that gas-phase pollutants played a substantial role in the observed toxic responses.

Statistical analysis strengthened that conclusion. Pearson’s correlation analysis revealed a strong relationship between LDH activity and ozone concentrations in both filtered and unfiltered exposures, with a double-logarithmic regression yielding a coefficient of determination of 0.33 and a p-value at or below 10 to the minus 6. Partial least squares regression, using log-transformed air quality parameters to predict toxicity, accounted for up to 80 percent of the variance in LDH values across three components. The dominant component alone explained 55.5 percent of the variance and was driven by a strong positive contribution from ozone and a negative contribution from black carbon, mirroring the inverse correlation observed between ozone and locally emitted combustion pollutants. Additional components linked residual toxicity to particulate matter, nitric oxide, and sulfur dioxide.

The authors are careful about what these associations mean. They note that correlation does not imply causation, and that ozone may be acting as a marker for other secondary pollutants formed through photochemistry, such as oxygenated volatile organic compounds, carbonyls, and secondary organic aerosols, none of which were measured in this study. Only inorganic gases were monitored, which limits the ability to disentangle the direct effect of ozone from that of co-varying photochemical products. Still, the combination of direct correlation, partial correlation, and partial least squares analyses led the team to describe ozone as the primary contributor to the observed toxicity among the measured pollutants, with particulate matter and nitrogen oxides contributing as well.

The findings also carry a methodological message. Traditional toxicological studies may underestimate ozone toxicity because laboratory samples and filter extracts rarely contain realistic concentrations of secondary gases, and reactive gases are poorly preserved in collection systems. By exposing cells to the atmosphere as it actually exists, the ALI approach allowed ozone to display cytotoxic effects that conventional sampling would have missed. The researchers also point out that ozone of regional origin, diluted by nitrogen oxide titration and ozonolysis reactions in the denser urban core, may have softened the expected toxicity contrast between the harbor and the quieter background site, so that pollution appeared widespread across the city rather than concentrated at the emission source.

The study has acknowledged limitations. The two-hour exposure duration keeps the accumulated dose low, which is why the LDH release assay was used instead of less sensitive viability tests, and the short timeframe means the findings reflect acute cellular responses rather than effects of chronic exposure. The experiments used a single cell line in vitro, so the results speak to mechanisms of membrane damage in alveolar epithelial cells, not to health outcomes in people. The authors frame the work as a demonstration that synchronized ALI exposure and air quality monitoring can identify the specific components responsible for harmful effects in real urban atmospheres, rather than as a risk assessment for the population of Barcelona.

Nevertheless, the implications reach beyond methodology. The toxicity patterns reflected the influence of transport emissions, from ships, cars, and trucks at the harbor and from road traffic near the background site, and the authors argue that reducing these emissions, ultimately through a transition away from fuel combustion in transport, is necessary to achieve a clean and safe urban environment. They also emphasize that both gas-phase and particulate pollutants must be considered when evaluating the health risks of urban air, since the relative contributions of gases and particles may shift with climate, sunlight, and atmospheric chemistry. For a field that has long concentrated its toxicological attention on particles, the message from the streets of Barcelona is that the invisible gas phase deserves equal scrutiny.

Subject of Research: In situ assessment of air pollution toxicity in Barcelona using a mobile air-liquid interface exposure system with human lung cells

Article Title: Assessment of air pollution toxicity in urban environments in Barcelona, Spain, using a mobile air-liquid interface system

Article References: Amato, F., Pyambri, M., Ridolfo, S., Bedia, C., Ricarte, M., Karanasiou, A., Rodriguez, A., Querol, X., Alastuey, A., Faria, M., Mansilla, S., Introna, M., Juárez-Facio, A. T., Steimer, S. S., Portugal, J., Piña, B., & Elihn, K. (2026). Assessment of air pollution toxicity in urban environments in Barcelona, Spain, using a mobile air-liquid interface system. Environmental Science and Pollution Research. https://doi.org/10.1007/s11356-026-38306-6

Image Credits: AI Generated

DOI: 10.1007/s11356-026-38306-6

Keywords: air pollution, ozone, particulate matter, air-liquid interface, A549 cells, cytotoxicity, Barcelona, harbor emissions, LDH assay, PLS regression, urban air quality, secondary pollutants