Mice. All animal procedures in this study were performed in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the Institute of Microbiology, Chinese Academy of Sciences (IMCAS) Ethics Committee.
Mice with less differences in body weight were randomly assigned to experimental groups, and no mice were excluded from the analysis. The 146 female mice used for this study are described in Supplementary Table 2.
8-week-old female ApoE−/− mice were fed with a high-fat diet (60 kcal% fat, 20 kcal% protein and 20 kcal% carbohydrate, Cat. D12492i, Research Diet, New Brunswick, NJ, USA) for 8 weeks, while 8-week-old female C57BL/6J as the background were fed on a normal chow diet (Lab Diet, Cat. 5001). All mice were housed in a specific pathogen-free (SPF) facility, with a strict 12 h light/dark cycle (8:00 a.m. to 8:00 p.m.), humidity at 50 ± 15%, a temperature of 22 ± 1 ℃, and ad libitum access to food and water. The health status of mice was determined via daily observation by technicians supported by veterinary care. All animal procedures in this study were performed in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the Institute of Microbiology, Chinese Academy of Sciences (IMCAS) Ethics Committee.
Mice with less differences in body weight were randomly assigned to experimental groups, and no mice were excluded from the analysis. The body weight of the mice was measured by weighing the mice on a scale once three days, and the body fat content and fat ratio were measured and calculated with an MRI body fat measurement instrument (EchoMRI-700, EchoMRI) two days before sacrifice.
Animals in each group were subjected to the OGTT and an ITT test before cervical dislocation sacrifice. The blood was sampled from the portal and cava veins and plasma was collected to measure BCAAs levels and other biochemical indicators. The liver, fat, muscle, feces, cecum, and intestines were quickly frozen and stored at -80 ℃, the whole aorta was harvested after heart perfusion and preserved in 4% paraformaldehyde. All efforts were made to minimize suffering.
In assays with GMD, HFD-fed ApoE−/− mice were sorted into four groups, GMD-treated groups were given with GMD (10 mg/kg and 5 mg/kg) daily by gavage. Atorvastatin-treated groups were administered 10.0 mg/kg daily by gavage. Treatments were continued for 8 weeks.
In assays involving live P. merdae and heat-killed P. merdae (KPM), HFD-fed ApoE−/− mice were sorted into three groups. Mice in the PM group were treated daily with 2 × 108 cfu of live P. merdae in 0.2 mL of sterile anaerobic PBS by mouth. Mice in the KPM group were given 2 × 108 cfu of heat-killed (Pasteurization consisted of heat treatment at 70°C for 30 min of fresh P. merdae) P. merdae suspended in 0.2 mL of sterile anaerobic PBS daily. The vehicle group was given an equivalent volume of sterile anaerobic PBS. Treatments were continued for 4 weeks. In assays with live P. merdae and P. merdae ΔPorA, HFD-fed ApoE−/− mice were sorted into three groups and using the same method as described for the PM and KPM experiments.
Microbial strains. The P. merdae strain was isolated from human feces by the microfluidic streak plate (MSP) method.59 The strain was identified by comparing 16S rRNA gene sequence with those in the NCBI reference database (https://www.ncbi.nlm.nih.gov/). The P. merdae strain was cultured in yeast extract, casitone and fatty acid (YCFA) medium at 37°C in an anaerobic chamber for 24 hours. For the in vivo efficacy assay, cell pellets of the wild-type and mutant strains of P. merdae were obtained by centrifuging at 8,000 × g for 10 min at 4°C. The cell suspension for oral administration was prepared by suspending the cultured bacterial cells in oxygen-free PBS at a final cell density of 1×109 CFU/mL and 200 µL of the bacterial suspension was given daily by gavage. The heat-killed P. merdae control was prepared by heating a culture at 70°C for 30 min.
Imaging of arterial lesions with high frequency ultrasound. A high frequency ultrasound system (Vevo 2100, Visualsonics, Toronto, Canada) equipped with a linear array transducer (MS 550D, 22–55 MHz) was used to detect atherosclerotic lesions at the aortic sinus. Briefly, ApoE−/− mice anesthetized with isoflurane were placed on a heated procedural board and limbs were taped to electrocardiogram electrodes coated with electrode cream. A rectal thermometer was inserted to assist with maintaining normothermia (37°C). The fur at the imaging location was shaved and warm ultrasound gel was liberally applied to ensure optimal image quality. The aortic sinus was imaged and visualized using a long-axis view; a CINE loop of 100 frames was stored for later off-line analysis. The time gain compensation curve was adjusted to produce uniform echo intensity. The gain was set to 30 dB and the dynamic range to 65 dB. To reduce variability, image parameters were held constant throughout the experiment. All examinations were performed by an experienced operator, and measurements were repeated three times at each site.
Assessment of aortic atherosclerotic lesion areas. Aortic specimens were resected and fixed with 10% formaldehyde in phosphate-buffered saline (pH 7.4), embedded in paraffin, sectioned and stained with hematoxylin and eosin (H&E), oil red O, or Masson’s stain. For staining of atherosclerotic lesions in the entire aorta, the aorta-to-iliac arteries were dissected and opened along the ventral axis. The aortas were pinned onto black wax in a pan and stained with oil red O. Images of the aorta were captured with a digital camera equipped with a Canon EOS 650D lens and further analyzed using ImageJ software. For histological analysis of atherosclerotic lesions in the aortic root, sections including the aortic valve were stained with H&E, oil red O, or Masson’s stain. Quantitative analysis of lesion areas and lipid-stained lesions was performed using ImageJ software.
Fecal collection. Animals were kept in an empty cage without bedding for 15 min to gather fresh stool samples into tubes, which were stored at -80ºC until analysis.
Cecal DNA extraction, library preparation, and sequencing. Cecal DNA was extracted using the QIAamp DNA stool mini kit (Qiagen). The V3–V4 region of the bacterial 16S rRNA was amplified by polymerase chain reaction using 20 ng of DNA, Platinum Hot Start PCR master mix (Thermo Fisher Scientific), the primers, F: ACTCCTACGGGAGGCAGCA, and R: GGACTACHVGGGTWTCTAAT, and sequenced on the HiSeq PE250 platform (Illumina) using the 2 x 250 bp paired-end protocol.
16S rRNA gene bioinformatics analyses. The overall amplicon sequencing data were from two experiments: (1) ApoE−/− mice treated with GMD versus vehicle, and (2) ApoE−/− mice treated with live P. merdae versus vehicle (Supplementary Table 3). Using the same QIIME2 workflow, 16S data were analyzed (see Data Supplement for details). The obtained pair-end reads were trimmed and then assembled with PANDAseq. 60 After filtering the chimeras by USEARCH, sequences were clustered into operational taxonomic units (OTUs) at a similarity cutoff value of ≥97% using the UPARSE algorithm. A representative sequence of each OTU was assigned to the taxa at genus level in the optimized version of the RDP database (http://rdp.cme.msu.edu). Each unique OTU was subjected to BLAST against the NCBI 16S rRNA database to identify the closest match to the taxa at species level based on lowest e-value and identity 97%. Abundances were recovered by mapping the de-multiplexed reads to the UPARSE OTUs. A rarefied OTU table from the output files was further analyzed with a visualization toolkit. The resulting abundance table and taxonomic classification was loaded into R. Statistical analysis of differentially abundant sequences and taxa was performed using DESeq2 1.16.161 and the log2 fold changes (log2FC) were obtained from comparison with the reference level.
Computational identification of genes associated with BCAA degradation. To identify and quantify the abundance of porA genes across the publicly available human gut metagenomes, the quality-control of processed reads was assessed using FASTQC (FastQC, bioinformatics.babraham.ac.uk/projects/fastqc/). Specifically, we used HUMAnN 3.0 to create an abundance profile of microbial metabolic pathways based on UniRef90 sequences and annotations.62 Next we used Diamond BLASTp to identify the nine homologs of the porA amino acid sequence in the 405 human gut metagenomes.63 The nine query sequences consisted of experimentally characterized porA sequences identified through homology searches (Supplementary Table 1). A reference gene was considered a homolog if it was aligned with one of the six query genes with ≥50% amino acid identity (AAID) over ≥70% of each gene’s length. Next, we obtained the relative abundance of the porA homologs across the 405 gut microbiome samples from Chinese individuals.38
Relative abundance of P. merdae in human gut metagenomes. After quality control, the 405 human gut metagenomes were annotated by species with kraken2,64 using the kraken2_hGMB database for comparison .59 Next, we calculated the relative abundance of P. merdae across the gut microbiome samples.
Biochemical and immunological assays. Levels of plasma glucose, TC, TG, HDL-C, LDL-C, HbA1C, TNF-α, IL-1β, LPS, ox-LDL, hs-CRP, MCP-1, sIgA and insulin were measured by commercial kits. The insulin sensitivity index (ISI) was calculated from the values of fasting blood glucose (FBG, in mg/dL) and fasting blood insulin (FBI, in mU/L). ISI = 1/1000 (FBG×FBI). An ITT was performed by injecting insulin (0.6 U/kg) intraperitoneally after 6 h of fasting. An OGTT was performed by giving a glucose bolus (2 g/kg) by gavage after overnight fasting. The level of blood glucose was measured using a glucose meter (Accu-Chek, Roche, Switzerland) before oral glucose load (0 min) and at 40, 100, and 160 min after oral glucose load. The AUCs generated from the data collected during the ITT and OGTT were calculated with GraphPad 8.0.
For lipoprotein profiling, 100 µL of plasma was separated by fast-performance liquid chromatography (FPLC) on a Superose S-6 10/300 GL column (GE
Healthcare, Sweden) at a flow rate of 0.5 mL/min. Forty sequential fractions of 500 µL each were collected, and triglyceride concentration was measured in each fraction.
Immunofluorescence analysis. Tissue sections were blocked with 5% bovine serum albumin (BSA) for 1 h, and incubated with the antibodies sequentially. Primary antibodies (Goat anti-Mouse IgA Antibody HRP Conjugated, Bethyl, RRID: AB_67140; Rabbit Anti-Mouse Phospho-S6 Ribosomal Protein, Cell Signaling, RRID:AB_331679; Rat Anti Mouse CD68, AbD Serotec, RRID: AB_2074849) were used at 1:250 dilution. Species-specific fluorescent secondary antibodies [Donkey Anti-Rabbit IgG H&L (Alexa Fluor® 647), Abcam, RRID: AB_2752244; Donkey Anti-Rat IgG H&L (Alexa Fluor® 488), Abcam, RRID: AB_2737355; Donkey Anti-Goat IgG H&L (Alexa Fluor® 594), Abcam, RRID: AB_2810222] were used at 1:250 dilution. An LSM 700 confocal microscope (Zeiss) was used for image acquisition; images were quantified with the Zen microscope software 2011, blue edition (Zeiss).
Real-time qPCR Analysis. Total RNA was extracted and purified from liver, fat, and muscle tissue following the protocol described in the blood and tissue kit with TRIzol reagent and RNAeasy Mini Kit (Qiagen). Quantification and integrity analysis of total RNA were performed by running 1 µL of each sample on an Agilent 2100 Bioanalyzer (Agilent RNA 6000 Nano Kit, Agilent). The cDNA was prepared by reverse transcription. Real-time qPCR was conducted with Gapdh mRNA as a housekeeping gene. Sequences of the primers used for real-time qPCR are shown in the supporting information (Supplementary Table 4). The qPCR mixture contained 100 ng of cDNA, 0.5 µM primers/0.15 µM probe, and Kapa Fast qPCR Mix (Kapa Biosystems). PCR amplification was performed using the following cycling parameters: 3 min at 95°C, 37 cycles of 3 s at 95°C, and 30 s at 60°C. The copy number was determined from the standard curve generated using a synthetic template.
Generation of a porA gene-deficient P. merdae strain. An internal fragment (615 bp) of the porA gene was cloned into the pGERM suicide vector incorporating E. coli (bla) and P. merdae (ermG) selective markers. The resulting construct was transformed into the conjugative E. coli WM3064 strain, which is auxotrophic for diaminopimelic acid (DAP). The E. coli donor strain WM3064 was grown aerobically at 37°C in Luria broth (LB) supplemented with DAP (100 µg/ml) and ampicillin (50 µg/ml) until it reached an OD600 of 0.2. The P. merdae recipient was grown anaerobically at 37°C in YCFA medium until it reached an OD600 of 1.0. A 2 ml mixture of equal volumes of donor and recipient cultures (1:1 ratio) was centrifuged, the supernatant was discarded and the mating mix was placed on a YCFA-medium agar plate. After aerobic overnight incubation, the plate was transferred into an anaerobic station and the bacteria were suspended in 5 ml of gifu anaerobic medium. After 5 h anaerobic incubation at 37°C, 100 µL of cell suspension was spread on YCFA medium agar plates supplemented (mutant selection) or not (control) with 25 µg/mL erythromycin. After four days of anaerobic incubation, erythromycin-resistant colonies were picked and used for genomic DNA extraction. Plasmid insertion into the target gene was then verified by PCR using primers targeting junction regions between pGERM and PorA gene (Supplementary Table 4).
High-performance gas chromatography. The extractions of BCAAs and SCFAs were performed at 4°C. One mL of 5 mM NaOH containing an internal standard (5µg/mL [2H3]-L-leucine, Sigma-Aldrich) was added to fecal samples (50-100 mg), and the samples were homogenized for 10 min and centrifuged at 12,000 r at 4°C for 20 min. Aliquots of 600 µL of fecal homogenate were transferred into 10 mL glass centrifuge tubes, and 200 µL of sterile DIW was added. For plasma samples, 100 µL of each sample and 500 µL of 5 mM NaOH containing internal standard were mixed in a 10 mL glass centrifuge tube. An aliquot of 500 µL propanol/pyridine mixture solvent (v/v = 3:2) and 100 µL of propyl chloroformate were subsequently added to the glass tube and vortexed briefly. The derivatization reaction was continued with ultrasonication for 1 min. The derivatives were extracted twice with hexane and anhydrous sodium sulfate (~10 mg) was added to remove traces of water. GC-MS analysis was performed using an Agilent 7890A gas chromatography system.
DNA isolation and qPCR. Bacterial DNA was extracted from feces of mice with the QIAamp DNA stool mini kit (Qiagen), qPCR was performed on a 7500 Fast Real-Time PCR System using the primers (Supplementary Table 4).
Antibiotic treatment. Ampicillin (1mg/mL), vancomycin (5 mg/mL), neomycin (10 mg/mL) and metronidazole (10 mg/mL) were provided in drinking water for two weeks to specific-pathogen-free (SPF) ApoE−/− mice. Antibiotic-treated mice were maintained in sterile cages, given sterile food and water, and handled aseptically.
Quantification and statistical analysis. GraphPad Prism version 8.0 (GraphPad Software, San Diego, CA) was used for statistical analysis. The experimental data are shown as the mean ± s.e.m. The sample size was estimated on the basis of previous experience, sample availability and previously reported studies. The numbers per group in the figure legends refer to the number of mice per group. We collected data from animal studies in a blinded manner, and no data were excluded from the analysis. The normal distribution of the data was determined by the D'Agostino and Pearson omnibus normality test. For statistical comparisons, Student’s t-test or one-way ANOVA with Tukey’s test was used to compare normally distributed variables. Non-normally distributed data were compared by the Mann-Whitney U test (between two groups) or the Kruskal-Wallis test (among multiple groups). Spearman’s correlations between changes in microbial species and host BCAAs and BSCFAs levels were calculated based on species with significant differences between the two groups. The Benjamini-Hochberg procedure with a cutoff of 0.1 was applied to all Spearman’s correlations. P < 0.05 was considered significant.