Ambient Ultrafine Particle Ingestion Alters Gut Microbiota in Association with Increased Atherogenic Lipid Metabolites

Ambient particulate matter (PM) exposure is associated with atherosclerosis and inflammatory bowel disease. Ultrafine particles (UFP, dp < 0.1–0.2 μm) are redox active components of PM. We hypothesized that orally ingested UFP promoted atherogenic lipid metabolites in both the intestine and plasma via altered gut microbiota composition. Low density lipoprotein receptor-null (Ldlr−/−) mice on a high-fat diet were orally administered with vehicle control or UFP (40 μg/mouse/day) for 3 days a week. After 10 weeks, UFP ingested mice developed macrophage and neutrophil infiltration in the intestinal villi, accompanied by elevated cholesterol but reduced coprostanol levels in the cecum, as well as elevated atherogenic lysophosphatidylcholine (LPC 18:1) and lysophosphatidic acids (LPAs) in the intestine and plasma. At the phylum level, Principle Component Analysis revealed significant segregation of microbiota compositions which was validated by Beta diversity analysis. UFP-exposed mice developed increased abundance in Verrocomicrobia but decreased Actinobacteria, Cyanobacteria, and Firmicutes as well as a reduced diversity in microbiome. Spearman’s analysis negatively correlated Actinobacteria with cecal cholesterol, intestinal and plasma LPC18:1, and Firmicutes and Cyanobacteria with plasma LPC 18:1. Thus, ultrafine particles ingestion alters gut microbiota composition, accompanied by increased atherogenic lipid metabolites. These findings implicate the gut-vascular axis in a atherosclerosis model.


16S ribosomal RNA sequence analysis
Raw MiSeq sequence data was processed in QIIME 1.9.1 using default quality parameters (Caporaso et al. 2010). Closed-reference operational taxonomic unit (OTU) picking at the 97% similarity threshold was performed in QIIME against the May 2013 version of the Greengenes database (http://greengenes.secondgenome.com), pre-filtered at 97% identity. Alpha and beta diversity were assessed using data rarefied to 32,346 sequences. Principal coordinates analysis (PCoA) was performed with distance matrices calculated using unweighted and weighted UniFrac (Lozupone et al. 2005). Adonis with 100,000 permutations was used to assess statistical significance of differences in beta diversity (Anderson 2001).
Differential abundance testing of OTUs, genera, or phyla was performed using phyloseq and the DESeq2 algorithm (http://www.bioconductor.org/packages/release/bioc/html/DESeq2.html) (McMurdie et al. 2013, Love et al. 2014. OTUs present in only a single sample were removed prior to analysis. An empirical Bayesian approach was used to shrink dispersion of normalized count data. Log fold changes for each OTU were fitted to a negative binomial model. Taxa were filtered out by choosing a mean count threshold maximizing the number of taxa returned at a given false discovery rate. Outliers were replaced by trimmed means. P-values were calculated using the Wald test then converted to q-values (http://www.bioconductor.org/packages/release/bioc/html/qvalue.html).

Gut permeability assay
In vivo intestinal permeability assay: The assay is essentially done as previously described (Moussaoui et al. 2014). Briefly, mice were given an oral gavage of fluorescein isothiocyanate (FITC) labeled 4-kD dextran (FD4, Sigma)) at the dose of 12 mg/ml in 0.9% NaCl. After 4 hours, mice were anesthetized with isoflurane and lightly hand restrained and a blood sample (200 μl) was withdrawn within a few seconds from the facial vein. Samples were placed immediately in heparin-coated tubes on ice and centrifuged. Intestinal permeability was determined by measuring plasmatic FD4 concentration using an automatic synergy HT plate reader (Ex 485 nm; Em 525 nm, BioTek).
Ex-vivo permeability assay: At the end of 10 week UFP exposure, pieces of ileum tissue from euthanized mice were mounted on 0.10 cm 2 slides dedicated to Ussing chamber. The intestinal paracellular permeability were assessed by measuring mucosal-to-serosal flux of FITC-labeled 4-kDa dextran (FD4; 2.2 mg/ml; Sigma) across the intestinal strip every 30 min for 2 hours as done previously with a slight modification (Moussaoui et al. 2014).

In vitro permeability assay
Enterocytes CACO-2 were seeded into clear transwell (Corning Inc, 6.5mm diameter, 0.4µm pore size)at 1X10 5 per well in 24-well plate. Cells were grown to complete confluence in 7-10 days with media change every 2 days. Cells were then rinsed with treatment media (M199/0.1%FBS) and treated with or without UFP in treatment media in the presence of 1µg/ml of Horse Radish Peroxidase-Streptavidin (HRP-Streptavidin, Thermofisher) in the top transwell.
One ml of treatment media was added to the bottom well. Changes in endothelial permeability were analyzed by the HRP activities of media in the bottom well.
For the HRP activity assay, five l of media from bottom well were taken and mixed with 100 l of TMB solution (HRP substrate, Thermofisher Inc) in wells of 96-well plate. After incubation for 5-30 minutes at a room temperature, one hundred l of 2M sulfuric acid was added to stop the reaction. The absorbance at 450 nm (OD 450) was read as HRP activities.

References:
Anderson  Figure S2. UFP ingestion reduced microbial diversity compared to vehicle control (VC) in Ldlr-null mice, associated with expansion of Akkermansia muciniphila and depletion of microbes within the Cyanobacteria, Actinobacteria, and Firmicutes phyla. (A) Alpha diversity is compared between the UFP and VC groups using three metrics: Chao1 (richness), Faith's phylogenetic diversity, and Shannon index (evenness). P-values were calculated using two-tailed t-tests. (B) Principal Coordinates Analysis plots are shown to visualize differences in microbial composition across samples (beta diversity) as assessed by unweighted and weighted UniFrac. Each dot represents a mouse, colored by exposure status. P-values were calculated using Adonis. (C) OTUs with a statistically significant difference in abundance between the UFP and VC groups are shown. Effect size is represented as the log2 fold change (FC). The size of each dot is proportional to the normalized abundance of that OTU. The genus of each OTU is shown on the horizontal axis. Genera are ordered by the log2 FC between the UFP and VC groups of the genus as a whole. Families [f] or orders [o] are indicated for OTUs that did not have an assigned genus in the Greengenes database. Color signifies phylum. Q-values are shown for the significance of phylum level changes in abundance. Figure S3. UFP exposure via oral gavage did not alter plasma levels of cholesterol. The plasma levels of total cholesterol from mice exposed to vehicle control or UFP were measured with a colorimetric assay kit from Pointe Scientific. There was no significant change of plasma cholesterol.