Edinburgh Napier University

  Particulate air pollution has been consistently linked with an increase in morbidity and mortality due to respiratory complications and cardiovascular disease. The aim of this study was to investigate which components of PM10 are important in driving its biological effects. In vivo studies: PM10 was collected daily for one year onto Teflon filters at 6 sites in the UK, each having air pollution sources of different characteristics. Particulate was extracted from the filters, before being either entered into a rat lung instillation model (in vivo) or used to treat cells in culture (in vitro). Estimated daily concentrations of primary, secondary and coarse particles at each location were calculated by source apportionment modelling (AEA Technology). For 6 months of the collection period concurrent PM10 samples were collected and analysed for metal content by ICP-MS (Edinburgh University).

Each 24h PM10 sample was extracted into 1 ml of sterile saline and the concentration of particulate extract was estimated by turbidometry. There was a strong correlation between PM10 mass on filters and the mass extracted. PM10 suspensions were not equalised for mass prior to instillation, instead extracts were instilled neat in order to represent the dose of particulate at respective sites on the dates studied. Each PM10 sample (0.5 ml) was intratracheally instilled into the lungs of one rat on one occasion. After 18 hours the lungs were washed to obtain broncho-alveolar lavage (BAL) cells and fluid. Preparations of BAL cells were counted to assess the presence of different cell types, especially neutrophils as an indicator of inflammation. BAL fluid was analysed for various biochemical parameters to determine the extent of lung damage.

On taking an average of the data obtained for PM10 at each location, the percentage of neutrophils in BAL cells was raised in lungs exposed to PM10 from all sites, but most notably Marylebone Road and Belfast. The concentration of macrophage inflammatory protein-2 (MIP-2), lactate dehydrogenase (LDH) and protein in BAL fluid were not significantly different from the control suggesting that the inflammation measured was not induced by gross lung damage.

On considering each instillation result independently, inflammation was found to be highly influenced by mass dose of PM10 instilled. Stepwise regression analysis of the results on the basis of composition, however, highlighted the importance of other factors in driving inflammation. Statistical analysis of PM10 samples selected from the entire 12 month sampling period revealed that primary particulate was a strong factor in determining potency whereas secondary and coarse particles were not. Analysis of PM10 samples selected from the last 6 months of sampling, coinciding with the PM10 collection for metal determination, revealed zinc to be highly inflammogenic, overriding the primary component in its influence on potency.

In conclusion, the results of the in vivo study indicate that mass continues to be an appropriate metric by which to monitor air pollution and confirm that a reduction in particulate concentration reduces the potency risk. However these results suggest that mass alone is not the only driving force behind PM10 induced inflammation, and that the transition metal components, including zinc are of great importance. The data also suggests that decreasing the coarse and secondary components of PM10 are unlikely of to be of much benefit to public health.

PM10 and PM2.5 TEOM filters were obtained from the AUN archive. PM10 from three locations, with collection periods corresponding to dates of Teflon PM10 samples tested, was extracted and diluted to the mean concentration instilled for each location in the main in vivo study. For Marylebone Road, PM10 extracted from Teflon filters induced a significantly greater inflammatory reaction in the rat lung than corresponding material from TEOM filters. For North Kensington and Belfast locations there was no difference in potency between Teflon (non-heated) and TEOM (heated) PM10. This data suggests that at some locations such as busy roadside or kerbside locations heating of the PM10 sample by the TEOM may lead to loss of volatile components with the potential to drive inflammation in vivo. However, further investigation is required to confirm such a conclusion.

TEOM PM2.5 and TEOM PM10 from Marylebone Road and from the same sampling period was compared for potency on an equal mass basis in the instillation model. PM10 consistently induced a greater inflammatory response than PM2.5.

In vitro studies: Lung epithelial and macrophage cell lines were treated in culture with PM10 suspensions for 4 and 18 hours. Release of the cytokines interleukin-8 (IL-8) and tumour necrosis factor- (TNF-) (respectively) into the supernatant media was used as the marker of a pro-inflammatory effect. However, IL-8 protein released from epithelial cells was adsorbed onto particle surfaces, especially onto Marylebone Road and Birmingham PM10 and hence was rendered unreliable as a marker. Biochemical analysis indicated that PM10 exerted toxicity to epithelial cells resulting in release of lactate dehydrogenase (LDH) and a slight depletion in cellular glutathione (GSH) and adenosine triphosphate (ATP). As an alternative inflammatory marker, IL-8 gene expression (mRNA) was determined. In both bronchial and alveolar epithelial cells lines PM10, notably from Birmingham, Marylebone Road and Belfast, up-regulated IL-8 mRNA expression compared to that in control cells.

PM10 induced a strong inflammatory response in macrophages. TNF- was released in response to treatment of particulate from all locations. Decreased TNF- release upon treatment with Marylebone Road and Birmingham PM10 was shown to be due to increased cellular toxicity leading to reduced capacity to generate cytokine, rather than adsorption of PM10.

This study highlighted the difference in sensitivity to PM10 exposure of macrophages compared to epithelial cells. PM10 from all locations strongly stimulated TNF- production and release from macrophages. However, notable toxicity was induced by 4 hours of PM10 treatment of the macrophages in vitro, to the extent that the cells’ capacity to generate TNF- was severely impaired in response to Birmingham and Marylebone Road particulate. Toxicity was demonstrated by very high LDH release and marked depletion in GSH and ATP.

In common with the in vivo study there was particularly strong correlation between markers of inflammation and cellular toxicity and the water soluble metal content of PM10 used to treat macrophages. TNF- was negatively correlated with metal content due to the toxicity mentioned above, as was GSH and ATP, whereas increasing LDH release was strongly correlated with total combined water soluble metal concentration. Of the individual metal analysed, manganese was most strongly associated with the biochemical markers followed by zinc.

In conclusion this project has confirmed the importance of mass in driving the inflammatory effects of PM10 and suggests that mass remains an appropriate metric by which to measure air pollution and implement reduction strategies. Furthermore this study has highlighted the benefit of reducing certain contributing components including primary particulate and zinc