Quantitative proteomic analysis

For proteomic analysis, 3 x 10⁵ HTR-8/SVneo cells were seeded in T-75 culture flasks (Nunc, ThermoFisher Scientific) and exposed to Prague PM (50 ng/ml) once (n = 3, Acute) or daily (n = 3, Chronic) and cells harvested after 48 hours (single dose without media change) or 7 days (both daily doses and media changes), respectively, and compared to unexposed controls (n = 4) receiving vehicle only with media changes daily for the same periods. Cells were washed 4 times in PBS, before removal with trypsin, pelleted and stored at -80°C for analysis.

Proteomics on cell pellets was performed at the Proteomics Core Facility at Sahlgrenska Academy, University of Gothenburg (http://proteomics.cf.gu.se/proteomics). A Tandem Mass Tag (TMT) set containing ten isobaric tags or labels (10-plex) allowed comparison of ten samples in one experiment for analysis by Mass Spectrometry (MS). In total, 100 μg protein per sample was trypsin digested into peptides and uniquely tagged at the N-terminals before analysis by nano-Liquid Chromatography-MS/MS. Proteins were identified in MS-raw data using Proteome Discoverer version 1.4 (ThermoFisher Scientific) against the Human Swissprot Database version March 2017 (Swiss Institute of Bioinformatics, Switzerland) and relative quantification was based on the ratio of reporter ion intensities/isobaric labelling). The low variability and the high sensitivity of the methodology used at the Proteomics Core Facility, University of Gothenburg, allows fold change of ≥1.2, if p-value is <0.05, to be considered significant. Student’s t-test was used to determine significant fold changes compared to unexposed controls, and together with Ingenuity Pathway Analysis (IPA, Qiagen) and Reactome Analysis (https://reactome.org/) to determine the biological relationships. Heatmaps were generated by Morpheus, https://software.broadinstitute.org/morpheus) for differentially expressed proteins (at p < 0.05 level of significance, FC > 1.2), and IPA analysis was used to generate diagrammatic representation of significant pathways and networks affected by PM exposure.

Reverse Phase Protein Arrays (RPPA) were performed at the Victorian Centre for Functional Genomics (VCFG), and the ACRF Translational RPPA platform at Peter MacCallum Cancer Centre Melbourne, for proteomic validation studies as previously described [45]. Briefly, snap frozen protein lysates of PM exposed and control cells (n = 3 samples per group) were homogenised in CLB1 buffer (Zeptosens, Bayer), and quantified. Using a Caliper ALH3000 liquid handling robot (Perkin Elmer), samples were serially diluted before spotting onto ZeptoChips (Zeptosens) in duplicate using a Nano-Plotter-NP2.1 non-contact arrayer (GeSim). Chips were blocked, incubated with pre-validated primary antibodies diluted 1:500 for 20 hours before applying Alexa Fluor647 anti-rabbit secondary antibody (1:1000, 4 hours) (#Z-25308, ThermoFisher Scientific). Zeptosens instrument and software (version 3.1) were used to scan and calculate the relative fluorescence intensity. All samples were normalised to the background values reported in the secondary antibody-only negative controls. The antibodies used in the array, have been experimentally verified and highly-specific for signalling pathway components but include only a limited selection of pathway proteins in phosphorylated and non-phosphorylated form.