Collagen-induced arthritis.
Male 6-week-old DBA/1j mice were injected intradermally at the base of the tail on days 0 and 21 with 100µl of an emulsion containing 200µg nasal bovine type II collagen (CII, Elastin) in 0.01M glacial acetic acid and an equal volume of complete Freund’s adjuvant (Millipore Sigma). On day 21, mice were either left untreated or treated with 1% (vol/vol) DMB (TCI Chemicals) or 1% (vol/vol) DMBut (Sigma-Aldrich) or no additional additive (vehicle), in drinking water with 100mg/ml grape-flavored sugar-sweetened Kool-Aid (Kraft Foods) added to encourage consumption, or 100mg/kg FMC (Jubilant Biosys Limited) gavaged orally every other day. Mice were monitored for onset of arthritis and severity of disease until 35 days after the initial immunization. Disease for each of the four paws was scored on a scale from 0–4 according to established metrics: 0—no erythema or joint swelling; 1—erythema and one swollen digit; 2—erythema and two swollen digits; 3—erythema and three swollen digits; 4—ankylosis. The score for each paw was summed to generate a total score per mouse. The incidence of arthritis was defined as a non-zero score. Therefore, the rate of incidence, as a percentage, indicates how many mice in the treatment group had a non-zero score on a given day. At euthanasia on day 35, blood was collected by cardiac puncture and the serum stored at -20°C for future analysis. Spleens and inguinal lymph nodes were harvested and processed. Feces were collected at day 35 for 16S ribosomal RNA sequencing analysis. Terminal euthanasia of animals involved intraperitoneal injection of a mixture of ketamine (100mg/kg) and xylazine (10mg/kg), with subsequent cardiac puncture as a secondary method of euthanasia.
Cytokine quantification.
Serum taken from mice with CIA on day 35 post-immunization was analyzed for cytokine concentrations using a Meso Scale Discovery U-Plex assay platform according to manufacturer instructions. Assay plates were imaged on a MESO QuickPlex SQ 120 at the University of Colorado, Anschutz Medical Campus Human Immune Monitoring Shared Resource (HIMSR).
Overnight cell cultures treated with vehicle, DMB, and DMBut with or without the presence of stimulating agents were pelleted at 1660rpm for 5 minutes to collect cell-free supernatants for quantification of secreted cytokine concentrations via ELISA. Concentrations of secreted murine IL-6, IL-17A, IFN-γ, IL-1β, and TNFα were determined using DuoSet ELISA kits (R&D Systems) according to manufacturer instructions at room temperature. Briefly, assay plates were coated with capture antibody diluted to the appropriate working concentrations in PBS overnight and subsequently blocked with 1% bovine serum albumin (BSA) in PBS for 2 hours. Unknown supernatant samples were either undiluted (IL-1β), diluted 1:1 (IL-17A and IFN-γ), or diluted 1:10 (IL-6 and TNFα) in 1% BSA/PBS. Samples incubated on the assay plate for 2 hours. Assay plates were developed in 100µl TMB Substrate Solution (Thermofisher Scientific) for 20 minutes and subsequently stopped with 50µl of 2N H2SO4 stop solution. Assay plates were imaged immediately after addition of stop solution on a SpectraMax iD5 plate reader at 450nm with correction at 540nm.
Anti-collagen type II-IgG antibody quantification.
Serum from day 35 post-immunization was evaluated for anti-collagen type II antibody concentrations via enzyme-linked immunosorbent assay (ELISA). All steps were performed on ice. ELISA-grade CII (Chondrex) was diluted 1:10 in 1x collagen dilution buffer (Chondrex) and incubated on the assay plate at 4oC overnight with gentle rocking while covered with aluminum foil. The assay plate was blocked with 0.5% BSA (Sigma-Aldrich) in PBS for 4 hours at 4oC with gentle rocking. A relative standard was generated using serum from a mouse with robust CIA, not otherwise treated, diluted 1:1,000 in 0.5% BSA/PBS and serially diluted 1:4. Unknown serum samples were diluted 1:10,000 in 0.5% BSA/PBS and incubated on the assay plate overnight at 4oC with gentle rocking while covered in aluminum foil. Goat anti-mouse IgG Fab-HRP, IgG1-HRP, IgG2a-HRP, and IgG2b-HRP antibodies (Southern Biotech) were diluted 1:10,000 in PBS and incubated on the assay plate at room temperature for 2 hours with gentle shaking. The assay was developed with 100µl of 1:1 BD OptEIA TMB reagents (BD Bioscience) at room temperature for 20 minutes, and subsequently stopped with 100µl 2N H2SO4 stop solution. Assay plates were imaged immediately after addition of stop solution on a SpectraMax iD5 plate reader at 450nm with correction at 570nm.
Flow Cytometry.
Splenocytes were strained through 70µm cell strainers (Fisher Scientific) and washed with serum-free RPMI 1640. The cell suspensions were pelleted at 4oC and 300xg for 5 minutes and the supernatant was discarded. Lymphocytes were resuspended in 1ml 5% fetal bovine serum (FBS) in PBS. Splenocytes were resuspended in 1ml 1x red blood cell lysis buffer (Invitrogen) and incubated on ice for 5 minutes. Lysis was stopped with 10ml PBS and the cell suspension was pelleted at 4oC and 300xg for 5 minutes. Splenocytes were resuspended in 1ml 5% FBS in PBS. 100µl of each cell suspension was added to a 5ml polystyrene round-bottom tube (Corning) and incubated in 1µl Human TruStain FcX (Biolegend) for 5 minutes at 4oC. 10µl Brilliant Stain Buffer Plus (BD Biosciences) was added to each tube and cells were stained for viability and surface markers as noted in Supplemental Table 1. Stained cells incubated at 4oC for 30 minutes. Cells were washed with 1ml 5% FBS in PBS at 4oC and 300xg for 5 minutes, then fixed and made permeable in 1ml Foxp3/Transcription Factor Staining Buffer (Tonbo Biosciences). Cells incubated at 4oC for 30 minutes. Fixed cells were washed twice with 1ml 1x Flow Cytometry Perm Buffer (Tonbo Biosciences) at 4oC and 300xg for 5 minutes and stained for intracellular markers. Stained cells incubated at 4oC for 45 minutes, then washed with 1ml 1x Flow Cytometry Perm Buffer at 4oC and 300xg for 5 minutes and resuspended in 300µl 5% FBS in PBS for analysis. Analysis of data was performed using FlowJo (version 10.8.1). Supplemental Table 2 lists the definitions of the T lymphocyte populations presented.
TMA lyase inhibition assay.
Proteus mirabilis (ATCC 29906) was cultured in 5ml Difco Nutrient Broth (BD Biosciences) overnight at 37oC and 215 rpm without antibiotic selection. Overnight cultures were sub-cultured at a dilution of 1:20 in fresh nutrient broth and grown overnight at 37oC and 215 rpm to serve as the starting material for downstream assays.
Inhibition of the P. mirabilis TMA lyase enzyme complex CutC/CutD by DMB was assessed as previously described with some modifications [26, 27, 77]. Briefly, overnight P. mirabilis cultures were pelleted by centrifugation at 3000 rpm for 30 minutes and the broth supernatant was discarded. Cells were resuspended in 10 ml PBS and 400µl of cell suspension was allocated to 13x100mm screw cap culture tubes (Pyrex) with gas-tight 13mm-425 Mininert valve caps (Supelco). To determine functionality of endogenous P. mirabilis CutC/CutD, bacteria were incubated at 37oC in the presence of 0µM, 25µM, 50µM, 75µM, and 100µM D9-choline (Cambridge Isotope Laboratories) for 2, 4, 6, and 24 hours. To determine the inhibition of endogenous P. mirabilis CutC/CutD by DMB, bacteria were incubated in the presence of 1M, 10mM, 100µM, 1µM, 10nM, 100pM, and 1 pM DMB for 15 minutes, then 100µM D9-choline was added to the reaction vials. Reactions were performed at 37oC for 2 hours. Reactions were quenched with 200µl of cold 1M NaOH and submerged in a liquid nitrogen bath. 2ml hexanes, 1ml butanol, and 200µl 1N NaOH were added to the reaction vials, vortexed for 1 minute, and centrifuged for 15 minutes at 4oC and 2500rpm. The upper phase was transferred to a new 13x100mm screw cap culture tube with PTFE-lined caps and 200µl of 0.2N formic acid was added. Vials were vortexed for 1 minute and centrifuged for 15 minutes at 4oC and 2500rpm. The lower aqueous phase was collected and stored at -80oC until analysis by stable isotope dilution LC-MS/MS.
Bone marrow-derived macrophage differentiation.
Bone marrow-derived macrophages were isolated from 6-10-week-old DBA/1j mice as previously described [78]. Briefly, bone marrow was flushed and processed to a single cell suspension from the femur and tibia using 1ml cold PBS. The single cell suspension was cultured in 9ml of complete RPMI 1640 supplemented with 10% fetal bovine serum, 2% HEPES, 0.6% penicillin/streptomycin, 0.1% 2-mercaptoethanol, and 20ng/ml recombinant mouse GM-CSF (Peprotech). After 72 hours, the cell culture media was replaced with fresh differentiation media to remove non-adherent cells. After 6 days of culture, adherent cells were washed with cold PBS and resuspended in RPMI 1640 without recombinant mouse GM-CSF. Cells were stimulated with 10µg/µl ultrapure E. coli K12 lipopolysaccharide (InvivoGen) overnight.
Gas Chromatography.
Serum samples were thawed to room temperature. After a brief vortex, 50µL of each serum sample was transferred to a glass vial. 3N HCl (20 µL) was added followed by hexane (50 µL) and vortexed thoroughly. The material was then transferred to respective vial inserts (150 µL). All the samples were then centrifuged at 4°C, 10 min, 3000 rpm. The upper hexane layer was then taken out and transferred into respective new vial inserts and capped immediately for chromatography.
Frozen liver samples were ground in a mortar with pestle in liquid N2. The suspension of ground liver in liquid N2 was quickly poured into a pre-weighed glass vial, allowing the liquid N2 to evaporate prior to capping and weighing the sample. To each sample, hexane was added in a weight (mg) to volume (µL) ratio of 1:2 and stored at -20°C. 3N HCl (20 µL) was added to each hexane-suspension, sonicated for 15 min, and vortexed vigorously for 5 min. The 100 µL top liquid layer was then transferred to respective vial inserts and capped. All the samples were then centrifuged at 4°C for 10 min at 3000 rpm. The upper hexane layer (50 µL) was then transferred into a new vial insert and capped immediately for chromatography.
Hexane extracts (1 µL) were injected into a Trace 1310 GC coupled to a Thermo ISQ-LT MS, at splitless mode. The inlet was held at 250°C. Peak separation was achieved on a 30m DB-WAXUI column (J&W, 0.25 mm ID, 0.25 µm film thickness). Oven temperature was held at 80°C for 2 min, ramped at 20°C/min to 125°C, then ramped at 40°C/min to 175°C and then to 240°C at 20°C/min with a final hold for 20 min. Helium carrier gas flow was held at 1.2 mL/min. Temperatures of transfer line and ion source were both held at 250°C. SIM mode was used to scan ions m/z 57, 69, 87 for DMB and m/z 59, 57, 101 for DMBut with scan time of 0.1 sec/ion under electron impact mode. Peak integration was completed using Chromeleon software (ThermoFisher) (Supplemental Fig. 3b and 3c).
UHPLC-tandem mass spectrometry.
Frozen P. mirabilis broth and mouse serum samples were thawed on ice and extracted with ice cold methanol, acetonitrile, and water (5:3:2, respectively) at a 1:25 ratio. Frozen cecum and liver samples were weighed to the nearest 0.1mg and extracted at 15mg/ml in the same extraction buffer.
Extractions were vortexed for 30 minutes at 4oC and then insoluble materials were pelleted by centrifugation at 18,000xg for 10 minutes at 4oC. Supernatants were analyzed using a Thermo Vanquish UHPLC coupled to a Thermo Q Exactive MS and run in positive and negative ion modes (separate runs). Injection volumes were 20µl for serum and broth extracts and 10µl for tissue extracts. UHPLC phases were water (A) and acetonitrile (B) supplemented with 0.1% formic acid for positive mode runs and 1mM ammonium acetate for negative mode runs. Metabolites were separated on a Kinetex C18 column (2.1 x 150mm, 1.7µm, Phenomenex) equipped with a guard column using a 5-minute gradient method with the following conditions: Flow rate 0.45ml/min; column temperature 45oC; sample compartment temperature 7oC; solvent gradient: 0-0.5 minute 5% B, 0.5–1.1 minute 5–95% B, 1.1–2.75 minute hold at 95% B, 2.75-3 minutes 95 − 5% B, 3–5 minute hold at 5% B. The mass spectrometer was operated in full MS mode at a resolution of 70,000, maximum injection time of 200ms, microscans 2, automatic gain control (AGC) ions, electrospray source voltage 4.0 kV, capillary temperature 320oC, and sheath gas 45, auxiliary gas 25, and sweep gas 0 (all nitrogen). Instrument stability and quality control were assessed using replicate injections of a technical mixture every 15 runs as previously described [79, 80]. Raw data files were converted to mzXML using RawConverter and metabolites were annotated and peaks integrated using Maven [81–83] in conjunction with the Kyoto Encyclopedia of Genes and Genomes (KEGG) database.
For D9-trimethylamine (TMA), TMA, and TMAO measurements, the mass spectrometer was operated as above with a scan range of 50–750 m/z. For 3,3-diemethylbutyrylcarnitine measurements, the mass spectrometer was operated as above with a scan range of 65–900 m/z.
Microbiome analysis.
Bacterial profiles were determined by broad-range amplification and sequence analysis of 16S rRNA genes following our previously described methods [8, 84, 85]. In brief, amplicons were generated using primers that target approximately 400 base pairs of the V3V4 variable region of the 16S rRNA gene. PCR products were normalized using a SequalPrepTM kit (Invitrogen, Carlsbad, CA), pooled, lyophilized, purified and concentrated using a DNA Clean and Concentrator Kit (Zymo, Irvine, CA). Pooled amplicons were quantified using Qubit Fluorometer 2.0 (Invitrogen, Carlsbad, CA). The pool was diluted to 4nM and denatured with 0.2 N NaOH at room temperature. The denatured DNA was diluted to 15pM and spiked with 25% of the Illumina PhiX control DNA prior to loading the sequencer. Illumina paired-end sequencing was performed on the Miseq platform with versions v2.4 of the Miseq Control Software and of MiSeq Reporter, using a 600-cycle version 3 reagent kit.
Illumina Miseq paired-end reads were aligned to human reference genome hg19 with bowtie2 and matching sequences discarded [86, 87]. As previously described, the remaining non-human paired-end sequences were sorted by sample via barcodes in the paired reads with a python script [88]. Sorted paired end sequence data were deposited in the NCBI Short Read Archive under accession number PRJNA1006768. The sorted paired reads were assembled using phrap [89, 90]. Pairs that did not assemble were discarded. Assembled sequence ends were trimmed over a moving window of 5 nucleotides until average quality met or exceeded 20. Trimmed sequences with more than 1 ambiguity or shorter than 350 nucleotides were discarded. Potential chimeras identified with Uchime (usearch6.0.203_i86linux32) using the Schloss Silva reference sequences were removed from subsequent analyses [91, 92]. Assembled sequences were aligned and classified with SINA (1.3.0-r23838) using the 418,497 bacterial sequences in Silva 115NR99 as reference configured to yield the Silva taxonomy [93, 94]. Operational taxonomic units (OTUs) were produced by clustering sequences with identical taxonomic assignments. This process generated 4136760 sequences for 23 samples (median sample size: 169668 sequences/sample; IQR: 113008 to 253290 sequences/sample). The median Goods coverage score was ≥ 99.97%. The software package Explicet (v2.10.5, www.explicet.org) was used for data organization and alpha-diversity calculations [95].
Statistics.
Unless specified otherwise, data was analyzed using GraphPad Prism software version 9; specific statistical tests for comparisons are referenced in the figure legends.
Ethics declarations.
All animal studies and methods were approved by the University of Colorado School of Medicine Institutional Animal Care and Use Committee (protocol #173). All animal studies and methods were performed in accordance with the ethics guidelines and regulations put forth by the University of Colorado School of Medicine Institutional Animal Care and Use Committee. All studies are reported in accordance with the ARRIVE 2.0 guidelines [96, 97].