Methods

Microbiome Glycine Lipid Extraction, High-Performance Liquid Chromatography Fractionation, and Liquid Chromatography-Mass Spectrometry Analysis

Microbiome GL-enriched fractions (L1256 and L654 class species) were extracted from the Bacteroidota member P gingivalis (ATCC 33277, type strain). While these lipids are found in gut Bacteroidota, our team has established robust purification methods using the oral bacterium, P gingivalis. Centrifuged bacterial samples were stored as lyophilized pellets until lipid extraction. Each bacterial sample was extracted using the procedure of Bligh and Dyer as modified by Garbus.18 The total lipids were fractionated by semipreparative high-performance liquid chromatography (HPLC) using neutral HPLC solvent, pooled, then refractionated using a normal phase HPLC column but with solvent supplemented with 0.1% acetic acid, as previously described.5

Fractions were evaluated by liquid chromatography-mass spectrometry (LC-MS) and fractions containing either L1256 or L654 species were pooled by lipid class, and sample purity was evaluated using LC-MS as previously described.14 LC-MS analysis of the L654- (Supplementary Figure 2) and L1256-enriched (Supplementary Figure 3) preparations confirmed they were composed predominantly of L654 and L1256 lipid species, respectively. The L654 preparation contained lower abundances of minor species, including L567, while the L1256 preparation contained some deesterified species. Analysis of lipid samples utilized a Sciex500 LC-MS instrument located in the Center for Environmental Sciences and Engineering at the University of Connecticut. The detection and quantification of serine/GLs and surrogate internal standard (D9-3-OH-C17:0) used negative ion ESI-MS/MS mode (multiple reaction monitoring [MRM]) as previously described.5,9 Each lipid class was identified based on characteristic retention times of purified or synthetic standards (see Supplementary Table 6), by either quantifying the characteristic molecular parent ion for each lipid class or by quantifying the dominant MS/MS transition ion characteristic of each serine/GL class (MRM) (Supplementary Table 7).

Sonicated lipid preparations were used in experiments, with vehicle CTL solutions undergoing the same sonication procedures as bacterial lipid solutions, as described previously.14

Animals and Diets

Male C57BL/6J and Ldlr−/− mice (aged 6 weeks) were obtained from Jackson Laboratory (Bar Harbor, ME) and allowed to acclimate for 2 weeks prior to the start of experiments. Three independent mouse studies were conducted.

In a diet intervention study, we determined the role of dietary fiber in modulating microbiome GL levels. Male Ldlr−/− mice (C57BL/6J background) were fed a standard low-fat chow diet (Inotiv Teklad 2918 irradiated diet) (Chow), a Western-type HFD at 5% w/w cellulose, or the same HFD but lacking cellulose and supplemented with fermentable fiber (HFD + fiber) for 8 weeks. Fiber supplementation consisted of a mixture of 5% w/w citrus peel pectin (mainly soluble fiber) (P9135, Sigma-Aldrich) and 5% w/w pea fiber (insoluble and soluble fiber) (HYPF-B3, Jianyuan) for a total of ∼7.5% w/w total dietary fiber and ∼5% w/w soluble fiber. Compositions of Western-type HFD and HFD + fiber are shown in Supplementary Table 1.

In a MASH efficacy study, the effects of chronic microbiome GL administration were examined in a mouse model of diet-induced MASH. Male C57BL/6J mice were fed a HF/HS/HC diet containing 35% w/w fat, 35% w/w sucrose, and 2% w/w cholesterol for 14 weeks to induce liver disease (Supplementary Table 8). Mice were fed HF/HS/HC diets for another 8 weeks while being intraperitoneally injected with bacterial lipid extracts enriched in L654 species (1 μg) (L654), L1256 species (1 μg) (L1256) or vehicle CTL every 48 h. The CTL group received injections of sonicated vehicle (saline with 0.03% v/v ethanol), while the other 2 groups received the injections of 1 μg of the respective sonicated GL preparations (L654 or L1256). This chronic dosing regimen was chosen based on previous research studies administering L654 to mice.8,14

In a GL immune response study, the effects of chronic L654 lipid class administration on T cells were studied in mice. Male C57BL/6J mice were fed for 6 weeks with one of the following dietary regimens: (1) a standard low-fat chow diet (Inotiv Teklad 2918 irradiated diet) (Chow); (2) the HF/HS/HC diet (HFD); (3) the HF/HS/HC diet with vehicle CTL injections 3 times weekly (HFD-Veh); or (4) the HF/HS/HC diet with intraperitoneal injections of bacterial lipid extracts enriched in L654 species (1 μg) administered 3 times weekly (HFD-L654).

Following their respective diets and treatments, the mice were fasted for 6 to 8 h and anesthetized with a ketamine/xylazine cocktail (100 and 10 mg/kg, respectively). Terminal blood was then collected via cardiac puncture for serum isolation, followed by euthanasia. Animals were perfused with sterile saline to clear any residual blood from tissues, which may interfere with analyses. Fecal samples were collected and snap-frozen before storage at −20 °C. Tissues were harvested, weighed, snap-frozen in liquid nitrogen and then stored at −80 °C. Saline-perfused liver sections (left lateral lobe) were fixed in 10% neutral-buffered formalin for 48 h. All mice were housed in a temperature-controlled room and maintained in a 14-h light/10-h dark cycle at the University of Connecticut-Storrs vivarium. All procedures proposed in this study have been approved by the Animal Care and Use Committee of the University of Connecticut-Storrs. All animal experiments were in accordance with the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health.19

Indirect Calorimetry and Body Composition

After 4 weeks of injections with either the respective vehicle CTL or bacterial lipid extract treatments, the diet-induced MASH mice were individually housed in the Oxymax Comprehensive Lab Animal Monitoring System (Columbus Instruments) to measure activity and energy expenditure, as previously described.20 Following a 48-h acclimation period in the Comprehensive Lab Animal Monitoring System, the mice received an additional injection of the respective CTL or treatments. Body composition was assessed using Echo-MRI (Echo Medical Systems) to determine total lean body mass, total fat mass, total body water, and free body water.

Portal Blood and Intestinal Lymph Collection

Prior to portal blood and mesenteric lymph collection, C57BL/6J mice (aged 8–12 weeks old) were provided standard chow and water ad libitum. Portal blood was collected using ethylenediaminetetraacetic acid-coated syringes from mice anesthetized with ketamine/xylazine. For mesenteric lymph collection, mice were placed under isoflurane anesthesia (induction at 5%, maintenance at 2%), and mesenteric lymph duct cannulas were surgically implanted. Lymph was collected on ice, as previously described.21 All surgical procedures were approved by the University of Pittsburgh Internal Animal Care and Use Committee and comply with the NIH Guide for the Care and Use of Laboratory Animals.

Liver and Splenic Lymphocyte Isolation and Flow Cytometry

Spleen cells were isolated, as previously described,22 and resuspended in staining medium (Hanks' Balanced Salt Solution/4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid containing 2% fetal calf serum). Hepatic lymphocytes were isolated from murine liver tissue through mechanical dissociation followed by density gradient centrifugation, as adapted from established protocols.23 At sacrifice, livers were perfused with sterile saline to eliminate circulating peripheral blood lymphocytes. Liver lymphocytes were isolated by mechanically dissociating excised tissue through a 100 μm mesh, followed by density-gradient enrichment using 35% Percoll/Heparin and subsequent red blood cell lysis with ACK buffer. The final suspension yielded >2.5 × 106 viable lymphocytes per liver.

Phenotypic characterization of lymphocyte populations was performed using multicolor flow cytometry. Cells were incubated with fluorescently conjugated monoclonal antibodies targeting CD3 (pan-T cell marker), CD4 (helper T cells), CD8 (cytotoxic T cells), CD44 (activation/memory marker), and CD69 (early activation antigen) (BioLegend), following manufacturer-recommended titration protocols validated against positive and negative CTLs. Stained cell suspensions were analyzed using a BD FACSymphony A5 SE flow cytometer at the UConn Health Center Flow Cytometry Core Facility. Data acquisition employed BD FACSDiva software (version 8.0.1) using a standardized gating strategy.

Serum Biochemical Analysis

Total serum cholesterol, high-density lipoprotein cholesterol, and TG were measured using assay kits from Wako Diagnostics (Richmond, VA). Alanine aminotransferase levels were quantified according to the manufacturer's instructions (Pointe Scientific, Inc; Canton, MI). Paraoxonase 1 lactonase activity was assessed in serum using previously described methods.24 Soluble TLR2 and C-C motif chemokine ligand 2 were measured in serum via indirect sandwich ELISAs (R&D Systems; Minneapolis, MN).

Liver RNA Isolation, Complementary DNA Synthesis, and Real-Time Quantitative Reverse Transcriptase Polymerase Chain Reaction

Total RNA was isolated from liver tissue using TRIzol (Life Technologies, Carlsbad, CA), treated with DNase I and reverse-transcribed into complementary DNA (cDNA) using the iScript cDNA Synthesis Kit (Bio-Rad, Hercules, CA). Real-time quantitative reverse transcriptase polymerase chain reaction was performed with iTaq Universal SYBR Green Supermix (Bio-Rad) on a CFX96 Real-Time PCR Detection System (Bio-Rad). Gene expression was normalized to the geometric mean of the reference genes Gapdh (glyceraldehyde 3-phosphate dehydrogenase) and Rplp0 (ribosomal protein, large, P0) using the 2−ΔΔCt method. Primer sequences are provided in Supplementary Table 9.

Liver Lipid Extraction and Tissue Histology

Liver lipids were extracted with a modified Folch method as previously described.14 Following extraction and solubilization in 1% Triton X-100, total cholesterol, free cholesterol, and TG were measured using enzymatic kits (Wako Diagnostics; Richmond, VA). Cholesteryl esters were also calculated as (total cholesterol – free cholesterol) × 1.67.

For liver histology, fresh tissue sections were fixed in 10% neutral buffered formalin before being processed and stained with hematoxylin and eosin at the Connecticut Veterinary Medical Diagnostic Laboratory in Storrs, CT, as previously reported.14 The histopathological scoring for lipid vacuolation (0–5), ballooning (0–5), portal inflammation (0–5), lobular inflammation (0–5), and fibrosis (0–5) were done by a American College of Veterinary Pathologists board-certified pathologist blinded to treatment using slightly modified techniques reported previously.25,26 Liver sections were also stained with Picrosirius Red, as previously described.27 The Picrosirius Red–positive area of the liver sections was calculated using ImageJ (National Institutes of Health) image analysis software. Representative photomicrographs were captured for each treatment group.

Microbiome Glycine Lipid Treatment of HepG2 Cells and Mitochondrial Functional Analysis

Human HepG2 cells were obtained from ATCC (Manassas, VA) and used as an in vitro model of hepatocytes. Cells were cultured in a humidified incubator at 37 °C and 5% CO2 and maintained in low-glucose Dulbecco's modified Eagle medium (1 g/L glucose) containing sodium pyruvate, 10% fetal bovine serum (Hyclone, Logan, UT), 2 mM L-glutamine, 100 U/mL penicillin, and 100 μg/mL streptomycin antibiotic (ThermoFisher Scientific, Waltham, MA). HepG2 cells were treated with bacterial lipid extract preparations enriched in L654 (0.654 μg/mL, 6.54 μg/mL) or L1256 (1.256 μg/mL, 12.56 μg/mL), corresponding to approximately equimolar concentrations (∼1 μM and ∼10 μM) of the dominant lipid species (ie, L654 and L1256, respectively), or vehicle CTL every 24 h for a total of 72 h and then were plated into 24-well plates in serum-free media for mitochondrial functional analysis. Cells were subjected to a Mito Stress test using an XFe24 Extracellular Flux Analyzer (Seahorse Biosciences, North Billerica, MA), as described previously.28 After completing the assay, total DNA was extracted from each well using NucleoSpin Tissue kit (Macherey-Nagel Inc) to normalize the results.

Mitochondrial DNA Copy Number

Total DNA was extracted from liver tissue and cells, and the mtDNA copy number was determined by quantitative PCR using a SYBR Green method on a Bio-Rad CFX96 Real-Time system. For this analysis, 16S ribosomal RNA, cytochrome b, and nicotinamide adenine dinucleotide dehydrogenase 1 were mtDNA markers, while 18S ribosomal RNA, hexokinase-2, and lipoprotein lipase served as nuclear reference genes. Gene primers for mtDNA analysis are provided in Supplementary Table 10.

Quantification of Microbiome Glycine Lipids in Mouse and Human Samples

Liquid nitrogen-preserved samples of human primary liver cancer (liver tumor) and paired adjacent nontumor liver tissues (liver nontumor) were obtained from the UConn Health Center Research Tissue Registry/Repository (Biorepository) (Farmington, CT). Mouse serum, liver tissue, and fecal samples were collected and snap-frozen in liquid nitrogen at the time of sacrifice. Internal standard (D9-3-OH-C17:0) was added to samples (serum, plasma, lymph, liver, feces) and lipids were extracted using a modified Bligh and Dyer method and analyzed by ultra-performance liquid chromatography-quadrupole time of flight LC-MS (Sciex500) and MRM as described above and previously.14

Fecal Bacteroidota Relative Abundance

DNA was isolated from homogenized mouse fecal pellets using QIAamp Fast DNA Stool Mini kit (Qiagen) according to the manufacturer’s instructions. Relative abundance of Bacteroidota was determined by quantitative PCR using 16S rRNA gene–targeted universal (926F and 1062R) and Bacteroidota-specific primers (Bac960F and Bac1100R) following the methods of Yang et al.29

Statistical Analysis

Statistical significance was determined using one-way analysis of variance with Fisher Least Squares Difference test for multiple comparisons. The Kruskal-Wallis nonparametric test was applied to nonnormally distributed data. Spearman rank correlation was used to assess relationships between variables. A P value < .05 was considered statistically significant. All statistical analyses were conducted using GraphPad Prism (version 8) software, and data are presented as mean ± standard error of the mean. All authors had access to the study data and had reviewed and approved the final manuscript.