Frequency of classical versus non-classical circulating monocytes is inversely altered in children highly exposed to PM2.5
A subset of fifty children (mean age of 8.02±0.60 years) from a previously characterized air pollution cohort (17 ) were selected based on exposure to the lowest (7.47±1.45 µg/m3, N=25) versus highest (21.33±4.61 µg/m3, N=25) ambient levels of PM2.5 (P <0.001) (Fig. 1A). Subjects were excluded if they had taken oral immune suppressants within 5 days of the blood draw, had a history of allergen immunotherapy within 1 year of the clinical visit, had a chronic disease other than asthma, or had an acute infection. We utilized CyTOF analysis of a 35-marker panel focused on identifying circulating monocyte changes (Supplementary Table 1). Cell populations other than monocytes were manually excluded (Supplementary Fig. 1 and Fig. 1B). Major monocyte subsets from children exposed to low versus high levels of PM2.5 were computationally visualized (Fig. 1C). Environmental exposure to high levels of PM2.5 was associated with significantly increased frequency of total (Fig. 1D, P =0.008) and classical (CD14+/CD16-; Fig. 1E,P =0.004) but not intermediate (CD14+/CD16+; Fig. 1F,P =0.071) monocytes. Conversely, the frequency of non-classical monocytes (CD14lo/-/CD16+) was significantly diminished in those highly exposed to PM2.5 (Fig. 1G, P =0.015). Elevation of classical and reduction of non-classical circulating monocytes in children highly exposed to PM2.5 was further confirmed by comparison of absolute number (Supplementary Fig. 2). Monocyte subsets did not differ between healthy and asthmatic children exposed to high levels of PM2.5. Therefore, these results suggest that the level of total and major monocyte subsets, classical and non-classical, could reflect the impact of exposure to the fine particulate pollutants. Importantly, monocytes have been found to be more heterogeneous than the major subsets indicate (18 ), leading us to take an unsupervised approach to identify additional subsets for our investigation of the impact of exposure to PM2.5 on circulating monocytes from healthy versus asthmatic children.
Distinct monocyte clusters are associated with exposure toPM2.5 levels in healthy versus asthmatic children
Using an unsupervised FlowSOM approach (19 ), we identified 8 monocyte clusters that were projected onto a two-dimensional Uniform Manifold Approximation and Projection (UMAP) plot (Fig. 2A-B). Also, we determined the proportion of each monocyte cluster per subject and found considerable interpersonal variation (Supplementary Fig. 3). UMAPs in Supplementary Fig. 4 represent the expression levels of individual markers within the predicted monocyte clusters.
We next addressed whether specific monocyte clusters display an association with ambient exposure to PM2.5. In accordance with our initial gating approach (Fig. 1), the frequencies of classical monocytes from clusters 1, 2, and 3 were significantly increased in subjects highly exposed to PM2.5. In these subjects, the frequency of cluster 1 [FcεR1a- CCR5-CD163dim CD123dimCLEC4D+ CD11b+CD14+], the major monocyte population, as well as cluster 3 [FcεR1ahi CCR5+CD163+ CD123+CLEC4D- CD11bdimCD14+] were particularly enhanced in children diagnosed with asthma compared to the healthy control group. The frequency of these monocytes was not associated with the disease status in subjects exposed to low levels of PM2.5 (Fig. 2C).
Similarly, frequency of cluster 4 [CD16+CD11chi CD11bdimCD64dim CD14dim/-], i.e., intermediate monocytes, was associated with asthma in subjects highly exposed to PM2.5. In contrast, no significant association was observed in those exposed to low levels of PM2.5 regardless of disease status (Fig. 2C). Among non-classical monocytes, the frequency of cluster 5 [CD16+ CX3CR1dimCD11b- CD14-] was negatively associated with PM2.5 exposure. However, decreased frequency of these monocytes in asthmatic individuals was not significant compared to a healthy control group with exposure to high levels of PM2.5. Interestingly, a higher frequency of cluster 5 was associated with asthma diagnosis in subjects exposed to low levels of PM2.5. Other clusters did not show a significant difference between individuals exposed to low versus high PM2.5 or to those with or without asthma (Fig. 2C).