Multiomic Signatures of Traffic-Related Air Pollution in London Reveal Potential Short-Term Perturbations in Gut Microbiome-Related Pathways

This randomized crossover study investigated the metabolic and mRNA alterations associated with exposure to high and low traffic-related air pollution (TRAP) in 50 participants who were either healthy or were diagnosed with chronic pulmonary obstructive disease (COPD) or ischemic heart disease (IHD). For the first time, this study combined transcriptomics and serum metabolomics measured in the same participants over multiple time points (2 h before, and 2 and 24 h after exposure) and over two contrasted exposure regimes to identify potential multiomic modifications linked to TRAP exposure. With a multivariate normal model, we identified 78 metabolic features and 53 mRNA features associated with at least one TRAP exposure. Nitrogen dioxide (NO2) emerged as the dominant pollutant, with 67 unique associated metabolomic features. Pathway analysis and annotation of metabolic features consistently indicated perturbations in the tryptophan metabolism associated with NO2 exposure, particularly in the gut-microbiome-associated indole pathway. Conditional multiomics networks revealed complex and intricate mechanisms associated with TRAP exposure, with some effects persisting 24 h after exposure. Our findings indicate that exposure to TRAP can alter important physiological mechanisms even after a short-term exposure of a 2 h walk. We describe for the first time a potential link between NO2 exposure and perturbation of the microbiome-related pathways.


Table of Contents
List of all 38 mRNA that were found as significantly associated to at least one TRAP in our MVN model with their corresponding gene symbol and Beta coefficients.List of all 78 metabolic features that were found as significantly associated to at least one TRAP in our MVN model, aOer Bonferroni correcPon of the number of metabolic features (N=6,040), with their corresponding Beta coefficients and AnnotaPons, and the percentage of missing values before imputaPon in the dataset (N=295 observaPons).Bold beta coefficients indicate the significant features.In case more than one feature was associated with the same metabolite (different adducts, in-source fragments), an asterisk aOer the metabolite name indicates the feature with highest intensity.
SI Figure S1.Schematic representation of the study population and exclusion criteria.SI Figure S2.Score plots of the first (x-axis) and second (y-axis) principal components from principal component analysis (PCA) of 300 metabolic profiles of the 50 subjects.SI Figure S3.Boxplots showing distributions of measured exposures in Oxford Street and Hyde Park.SI Figure S4.Manhattan plots showing the -log10 transformed p-values of metabolic features plotted against their retention time for each TRAP exposure.SI Figure S5.Calibration plot for consensus clustering of metabolomics.SI Figure S6.Scatter plot of molecular mass of 78 metabolic features plotted against their retention time (min).SI Figure S7.Volcano plots for transcriptomics.SI Figure S8.Calibration plots for metabolomics networks.SI Figure S9.Calibration plots for multi-omic networks.

SI Figure S2 :
Score plots of the first (x-axis) and second (y-axis) principal components from principal component analysis (PCA) applied to the 300 metabolic profiles collected for the 50 participants.The 5 outlying observations: 9721, 7623, 1511, 6011 and 10123 were excluded for subsequent analysis.SI FigureS4: Manhattan Plots illustrating the association between each metabolic feature and each individual TRAP using our univariate analysis.Each metabolic feature is represented by its retention time (minutes, X-axis) and its -log10 p-value (Y-axis).Results are presented for each TRAP exposure, separately and metabolic features associated with TRAP exposures at a Bonferroni significance level correcting for ENT=284 and 6,040 tests are represented in blue and red, respectively.Calibration plot for consensus clustering of the 78 metabolic features selected with the MVN model as significantly associated to TRAP exposure.The stability score was plotted against the number of clusters ranging from n= 2 to n= 77.The highest stability score was achieved at 63 clusters (red dotted line).SI FigureS8: Calibration plots for metabolomics networks, for the three time points, 2 h before walks (time point 1), 2 h (time point 2) and 24 h (time point 3) after walks.The colour denotes the stability score for each combination of the two hyperparameters: penalisation (λ) and threshold of selection proportion (π).
SI Table 1.Characteristics of the study population.SI Table 2. Description of measured exposures.SI Table 3. List of mRNAs significantly associated with at least one TRAP with existing gene names.SI Table 4. List of metabolic features significantly associated with at least one TRAP, their cluster membership, and annotation.SI Table 5. Annotation of metabolic features.

SI Table 1: Characteristics of the study population with full exposure data in the Oxford Street II study.
COPD: Chronic obstructive pulmonary disease, IHD: ischaemic heart disease.P-values from paired t-tests for continuous variables and chi-square tests for categorical variables were shown.

Table 2 : Description of exposures in the Oxford Street II study.
BC: Black carbon, PCNT: Total number of particles.P-values were calculated using paired t-test.

Table 5 .
Details for annotaPon of metabolic features, including ion species, mass-tocharge raPo (m/z) and retenPon Pme for each feature, the percentage of missing values before imputaPon in the dataset (N=295 observaPons), and the level of annotaPon confidence according to the MSI scale.In case more than one feature was associated with the same metabolite (different adducts, in-source fragments), an asterisk aOer the metabolite name indicates the feature with highest intensity.