Impacts of Aronia on Inflammation and the Gut Microbiome

The overall goal of this project is to determine the inflammation lowering impact of anthocyanin-rich Aronia berries. Inflammation is an underlying mechanism driving the development of several diseases. While an elevation in immune signals in the systemic circulation is commonly attributed to adipose tissue, inflammation is not present in all obese individuals. Adipose tissue must become inflamed, and the inflammation trigger may come from other sources. Microorganisms (microbiome), host tissues, and immune cells residing in the gastrointestinal tract (GIT) are a key source of pro-inflammatory signals that may cause the host organism to become inflamed. Anthocyanins are bioactive compounds with established anti-inflammatory and microbiome altering properties. We hypothesize that the GIT microbiome is a key determinant of host inflammation than can be manipulated by anthocyanins-rich berries to lower inflammation. We assembled a cohort of individuals, characterized their GIT microbiome and performed anthropometric measurements, basal measures of metabolism and metabolic health, and triglyceridemic, metabolomic, and inflammation responses to a high-fat meal challenge.

Anthropometrics. Measurements were collected from participants using the validated segmental multifrequency bioelectrical impedance analysis (SECA mBCA 515, Hamburg, Germany). Fat mass (%) and estimated visceral adipose (L) were used for analysis. High-Fat Meal Challenge. The high-fat meal contained salted butter (58.3 g, Tillamook) over 3 pieces of whole wheat toast (127.5 g; Wheat Montana). Total energy content of the meal was 714 kcal, with 43.1% from fat, with a macronutrient breakdown of 50 g fat, 54 g carbohydrate, and 12 g protein. Water was provided with the meal; caffeinated black tea was provided instead for participants who identified as habitual coffee consumers. Blood Sampling. Participants were instructed to avoid alcohol consumption and strenuous physical activity in the 24 hours before their visit and to complete an overnight fast (10 - 12 hours) before blood collection. Participant blood samples were collected by a certified nurse or physician in the morning before ingestion of the meal and hourly for 4 hours after meal ingestion, totaling five time points. Whole blood in serum separating tubes was allowed to clot for 15 minutes before centrifugation at 1200 RPM for 15 minutes with resulting serum aliquoted and stored at -80ºC until analysis. Determination of blood markers. Blood markers of metabolic syndrome were determined from whole blood run on Picollo Xpress Chemistry Analyzer lipid panels (Abaxis, Union City, USA). Serum insulin (INS) was determined using an insulin ELISA kit (MP Biomedicals, Solon, OH) performed according to manufacturer instructions. Cytokine measurement was performed using high-sensitivity multiplexing technology (Bio-Rad Bio-Plex 200 HTS) following procedures by Millipore (EMD Millipore Corporation, Billerica, USA). Classic systemic pro-inflammatory cytokines were measured and include granulocyte macrophage colony stimulating factor (GM-CSF), interleukin (IL)-1B, IL-6, tumor necrosis factor (TNF)-α. InterleukinI-17 and IL-23, both of which serve a pro-inflammatory and regulatory role in the gut mucosa, were also measured. Serum samples at each time point during the high-fat meal challenge were run in duplicate. Stool Sample Collection. Collection kits were provided and participants were asked to follow included instructions for the self-collection of a stool sample in the 24 hours before their blood collection visit. After initial collection into a sterile disposable commode, a small portion of the sample was transferred into a sterile Eppendorf tube and transported to researchers. Samples were prepared and aliquoted in an anaerobic chamber then frozen at -80ºC until analysis. Genomic DNA Extraction and Microbial Analysis. Extraction of bulk DNA from fecal samples was performed using Powersoil DNA Isolation Kit (Mo Bio Laboratories, Inc.) and bead beating. DNA was shipped overnight to the University of Michigan, Michigan Microbiome Project for Illumina MiSeq amplicon sequencing of the 16S rRNA V4 region. After DNA quantification, V4 amplicon libraries were generated with dual-index barcoded primers, then by library purification, pooling, and MiSeq paired-end sequencing. Raw sequencing reads were processed and curated using MOTHUR software (Version 1.35.1) following the MOTHUR standard operating procedure for the MiSeq platform39. In brief review, paired-end reads were assembled into contiguous sequences and screened for length and quality. The remaining contigs were aligned to the SILVA ribosomal RNA database (Release 132), a comprehensive collection of aligned rRNA sequences. Potentially chimeric sequences were identified and removed using the UCHIME algorithm in MOTHUR. Taxonomic classifications were assigned using the Bayesian classifier of the Ribosomal Database Project. Non-target reads were removed, and operational taxonomic units (OTUs) were assigned using VSEARCH distance-based clustering at the 97% similarity threshold. Alpha-, and β- diversity indices were generated using the vegan package in R40. An OTU-based data matrix was constructed for participants included in the ppTG phenotype. Metabolomic Analysis. Frozen serum samples were thawed and 20μL was placed in a clean tube. 80μL of HPLC grade methanol was added to the sample after which it was vortexed briefly and placed in a -80 C freezer for 2 hours. After two hours, the sample was centrifuged at 20,000g for 10 minutes. The metabolite supernatant was collected and concentrated in a Speed Vac to dryness while the protein pellet was discarded. Samples were then stored at -80 C until ready for LCMS analysis at which time they were reconstituted with 40μL of methanol:water (50:50) and placed in a clean mass spectrometry vial. Analysis was completed on an Agilent 6538 Q-TOF MS coupled to an Agilent 1290 UHPLC using a 130A, 1.7μm, 2.1mm X 10mm Acquity BEH-HILIC HPLC column. Samples were ionized via electrospray ionization and runs were completed in positive mode. Mobile phase A was 15mmol/L ammonium formate and mobile phase B was ACN using a 10-40% A gradient over 6 minutes. Flow was kept at 400µL/minute and the column compartment temperature was set at 30 C. MSMS analysis was completed using the same LC conditions while targeting specific ions using retention time and m/z values from previous MS runs. After LCMS analysis completion, raw data files were converted to .xml files using MSConvert. Data was then mined with mzMine using an intensity minimum value of 1,000 based on a visual inspection of the total ion chromatogram to remove noise. Blank samples were also ran and the resulting features were removed from the biological data if present at a ratio under 5:1 in the sample compared to the blank. Mined data was then input into MetaboAnalyst for statistical analysis. Tandem MS data was analyzed with Sirius software to identify features.