Microbial enzymes induce colitis by reactivating triclosan in the mouse gastrointestinal tract

Ethical statement

The animal experiments were conducted in accordance with protocols approved by the Institutional Animal Care and Use Committee of the University of Massachusetts (Amherst, MA) and Massachusetts Host-Microbiome Center at the Brigham and Women’s Hospital (Boston, MA). The analysis of the de-identified human urine and stool samples, which are from a previous human study (ClinicalTrials.gov identifier NCT01509976)¹⁵, were conducted in accordance with the protocol approved by the Institutional Review Board of Stanford University. All subjects provided informed consent for their specimens to be used for studies of the microbiome; samples had been deidentified by the time this specific project was undertaken.

Chemicals

Triclosan (TCS, 99% purity) was purchased from Alfa Aesar (Haverhill, MA). TCS-glucuronide (TCS-G, 95% purity) and TCS-sulfate (95% purity) were from Santa Cruz Biotechnology (Dallas, TX). Stable isotope-labeled triclosan (¹³C₁₂-TCS, 99% purity) was obtained from Cambridge Isotope Laboratories (Andover, MA).

Animal experiments

Animal experiment 1: LC-MS/MS profiling of TCS metabolism in mice

C57BL/6 male mice (age = 6 weeks) were purchased from Charles River and maintained in a specific pathogen-free animal facility. The mice were treated with a modified AIN-93G diet which contains 1, 10, or 80 ppm TCS for 4 weeks, then the mice were sacrificed to harvest tissues for LC-MS/MS analysis. The composition of the diet is casein (200 g/kg), l-cystine (3 g/kg), sucrose (100 g/kg), dyetrose (132 g/kg), cornstarch (397.486 g/kg), cellulose (50 g/kg), mineral mix #210025 (35 g/kg), vitamin mix #310025 (10 g/kg), choline bitartrate(2.5 g/kg), corn oil (70 g/kg), and vitamin A palmitate (0.016 g/kg)¹². The ingredients for the preparation of the diet, except corn oil, were purchased from Dyets Inc. (Bethlehem, PA). The commercial sample of corn oil (Mazola, ACH Food company) was purchased from a local market in Amherst, MA, and purified by a silicic acid-activated charcoal chromatography to remove any pre-existing lipid oxidation compounds, then the purified oil was fortified with 400 ppm tocopherols, flushed with N₂, and stored at −80 °C until use.

Animal experiment 2: effects of antibiotic suppression of gut microbiota on TCS colonic metabolism in mice

C57BL/6 male mice (age = 6 weeks) were given drinking water with or without an antibiotic cocktail (a mixture of 1.0 g/L ampicillin and 0.5 g/L neomycin) throughout the entire experiment. This antibiotic composition was used in previous studies by others^20,21 and us^22,23. After 5 days, the mice were treated with a modified AIN-93G diet which contains 80 ppm TCS (see diet composition in animal experiment 1 above). After another 4 weeks, the mice were sacrificed, and their tissues were collected for LC-MS/MS analysis.

Animal experiment 3: effects of antibiotic suppression of gut microbiota on the kinetics of TCS colonic metabolism in mice

C57BL/6 male mice (age = 6 weeks) were supplied with drinking water with or without the antibiotic cocktail for 7 days, then the mice were treated with a one-time oral gavage of 8 mg/kg TCS which was dissolved in polyethylene glycol 400 (PEG-400). At t = 4, 8, 12, and 24 h post the oral gavage, the mice were sacrificed to harvest tissues for analysis. The Area under the curve (AUC) (Fig. 2) was calculated using GraphPad Prism software, Version 9.1.2 (225) (https://www.graphpad.com/scientific-software/prism/) with the parameters as follows: the baseline is set as Y = 0 and the peaks that are less than 10% of the distance from minimum to maximum Y are ignored.

Animal experiment 4: comparison of TCS colonic metabolism in conventional mice vs. germ-free mice

Conventional or germ-free male mice, established on C57BL/6 or Swiss Webster background, were treated with a one-time oral gavage of 8 mg/kg TCS which was dissolved in PEG-400. At t = 4–8 h post the oral gavage, the mice were sacrificed to harvest tissues for LC-MS/MS analysis.

Animal experiment 5: effects of GUSi on colitis-enhancing effects of TCS in mice

C57BL/6 male mice were orally gavaged with a specific GUS inhibitor (GUSi) UNC10201652 (dose = 1 mg/kg) or vehicle (a mixed solvent of 1:9 DMSO and saline) every other day throughout the experiment, as described previously^29,30. After 3 days, the mice were treated with a modified AIN-93G diet which contains 80 ppm TCS or vehicle (PEG-400) until the end of the experiment. After another 3 weeks, the mice were stimulated with 2% DSS (molecular weight = 36–50 KDa, MP Biomedicals, Solon, OH) in drinking water for 6 days to induce colitis. At end of the experiment, the mice were sacrificed for analysis.

Animal experiment 6: effects of GUSi on gut inflammation and gut microbiota in mice

C57BL/6 male mice were orally gavaged with GUSi UNC10201652 (dose = 1 mg/kg) or vehicle (a mixed solvent of 1:9 DMSO and saline) every other day for 24 days (the same treatment scheme as in animal experiment 5). At end of the experiment, the mouse feces were collected and subjected to sequencing, and the mice were sacrificed for biochemical analysis.

Detection of TCS and its metabolites by LC-MS/MS

Mouse tissues and human stool were placed in homogenizer tubes with beads and 1 mL methanol, then homogenized using a bead-disruptor (OMNI International, Kennesaw, GA). Samples were centrifuged at 10,000 rpm for 3 min. The supernatant was collected and then centrifuged again at 14,000 rpm for 5 min. In total, 500 μL of the supernatant was then collected, and vacuum centrifuged to dryness. For bacterial broth (50 μL) and human urine (100 μL), each sample was combined with 1 mL methanol and placed on ice. After 10 min on ice, samples were centrifuged at 14,000 rpm for 5 min. 500 μL of the supernatant was then collected and vacuum centrifuged to dryness. Stable isotope-labeled ¹³C₁₂-TCS was used as the surrogate standard during the extraction. The extracts were re-dissolved in methanol with the amount that was proportional to sample weights or volumes, then centrifugated (14,000 rpm, 15 min, 4 °C) before the LC-MS/MS analysis.

TCS, TCS-G, and TCS-sulfate in the samples were quantified using a Thermo Scientific Dionex Ultimate 3000 ultrahigh performance liquid chromatography (UHPLC) system coupled with a TSQ Quantiva Triple Quadrupole Mass Spectrometer. ACQUITY UPLC C18 column (1.7-μm particles, 2.1 × 100 mm, Waters) was used for chromatographic separation. Data acquisition was performed by multiple reaction monitoring (MRM) in negative ionization mode. Details of the instrumental methods are provided in Supplementary Table S4. The data were analyzed using Xcalibur software (version 4.1, Thermo Fisher Scientific).

The spike recoveries of the three target compounds in the matrixes of mouse colon digesta were determined. The recoveries (%, mean ± SEM) were 101.6 ± 8.9 and 95.6 ± 3.4 for TCS, 91.6 ± 5.4 and 87.1 ± 5.9 for TCS-G, 95.1 ± 1.4 and 96.9 ± 6.0 for TCS-sulfate, based on two spiked levels of 2 pmol/mg and 10 pmol/mg, respectively (n = 3 replicates). No significant differences were found among these three compounds. Therefore, ¹³C₁₂-TCS was used for the signal correction of TCS, TCS-G, and TCS-sulfate, and it is a strategy for the absolute quantitation of analytes when internal standards are unavailable^12,45. For the quantification of TCS, TCS-G, and TCS-sulfate by LC-MS/MS in the different experiments, blank samples from the control group without TCS exposure were used as the matrixes for calibration curve standards. During the instrumental analysis, the matrix calibration curve was performed at the beginning and at the end of every sample batch. All reported concentrations were determined based on a standard curve with 7–10 data points.

Isolation of bacteria from mouse tissues and human stool samples

Mouse fecal tissues and human stool samples were collected, dissolved in sterile PBS with 0.05% l-cysteine, then centrifuged at 900× g for 5 min. The supernatant containing culturable bacteria was then fermented at 37 °C in MRS broth in an anaerobic cabinet (Whitley A35 anaerobic workstation, Don Whitley Scientific) under an atmosphere of 85% N₂, 10% CO₂, and 5% H₂. In addition, the remaining supernatant (~0.6 mL) containing culturable bacteria was then mixed with sterile 50% glycerol (0.3 mL) and stored at −80 °C as stock for future experiments.

Protein gene synthesis, expression, and purification

All genes were codon-optimized for E. coli expression, synthesized, and ligated into a pLIC-His vector by BioBasic. Genes were transformed into BL21-G E. coli competent cells. A 100 mL culture was grown overnight at 37 °C in LB broth with ampicillin (100 µg/mL) and shaking at 215 RPM. The following day, 50 mL overnight culture was added to 1.5 L of LB broth with ampicillin and ~40 µL Antifoam 204. For FMN-binding enzymes, 500 µM FMN was added to the culture flask. The culture was incubated at 37 °C and 215 RPM until it reached an OD of 0.6. The culture was then induced with 1-thio-β-d-galactopyranoside (100 µM) and incubated overnight at 18 °C.

Cells were pelleted at 4,500 × g for 20 min at 4 °C in a Sorvall (model RC-3B) swinging bucket centrifuge. Pellets were resuspended in 35 mL Buffer A (20 mM potassium phosphate, 50 mM imidazole, 500 mM NaCl, pH 7.4. For FMN-binding enzymes, buffer contained 50 µM FMN) with DNAse, lysozyme, and one EDTA-free protease inhibitor tablet (Roche). The resuspension was sonicated twice using 1 s pulses for 1.5 min and the resultant suspension was pelleted at 17,000 × g for 45 min in a Beckman Coulter J2-HC centrifuge. The supernatant was syringe-filtered using a 0.22-µm filter.

The filtrate was flowed over a 5-mL nickel-nitrilotriacetic acid HP column (GE Healthcare) using the Aktaexpress FPLC (Amersham Bioscience) and washed with Buffer A. Protein was eluted using a linear gradient of Buffer A to Buffer B (20 mM potassium phosphate, 500 mM imidazole, 500 mM NaCl, pH 7.4. For FMN-binding enzymes, buffer contained 50 µM FMN). Fractions containing the protein of interest were collected and applied to a HiLoad 16/60 Superdex 200 gel-filtration column (GE Life Sciences). Samples were eluted in S200 buffer (20 mM HEPES, 50 mM NaCl, pH 7.4). Fractions containing the protein of interest were analyzed via SDS-PAGE. Those with >95% purity were combined and concentrated to ~10 mg/mL using 50 kDa cutoff molecular weight centrifuge concentrators (EMD Millipore). Samples were snap-frozen using liquid nitrogen and stored at −80 °C.

Site-directed mutagenesis

All mutants were created using site-directed mutagenesis. Primers were synthesized by IDT Technologies (Supplementary Table S5). Mutant plasmids were sequenced by Eton Biosciences to confirm mutation incorporation. Mutant proteins were purified using the same purification protocol described above.

Fecal extract preparation, proteomics, and analysis

Fecal extracts were prepared and proteomic analysis was performed exactly as previously described³⁵. In total, 10 g of frozen human fecal sample was thawed and added to 25 mL extraction buffer (25 mM HEPES, 25 mM NaCl pH 6.5 with Roche Complete protease inhibitor tablet) and 0.5 g autoclaved garnet beads. The mixture was vortexed until homogenous and centrifuged at low speed (300 × g for 5 min at 4 °C). The supernatant was decanted. In all, 25 mL buffer was added to the pellet and the vortex and centrifugation steps repeated. The supernatants from these steps were then combined and centrifuged at low speed for two more cycles. The resultant supernatant was sonicated twice using 1 s pulses for 1.5 min and the lysate was centrifuged at 17,000 × g for 20 min in a Beckman Coulter J2-HC centrifuge. The decanted lysate was then washed with several exchanges of extraction buffer to remove metabolites and small molecules. Total protein concentration was quantitated using a Bradford assay. The fecal lysate was diluted to 1 mg/mL concentration and snap-frozen in small aliquots.

In all, 3.5 mg purified fecal extract was incubated with 10 µM biotin-activity-based probe complex in 500 µL extraction buffer with 1% dimethyl sulfoxide for 1 h at 37 °C. To quench, 125 µl 10% sodium dodecyl sulfate (SDS) was added and samples were heated to 95 °C for 5 min. Samples were cooled on ice and washed with extraction buffer containing 0.05% SDS three times by centrifugation for 5 min at 14,000 × g in 1.5 mL Amicon 10 K cutoff spin concentrators. After centrifugation, the total volume was brought to 1 mL using extraction buffer with 0.05% SDS. In total, 15 µL streptavidin sepharose beads (GE) were added and samples incubated at room temperature for 1 h. Beads were then washed 3 times with 300 µL extraction buffer with 0.1% SDS, three times with 300 µL extraction buffer alone, and three times with 300 µL 50 mM NH₄HCO₃. Samples were centrifuged at 400 × g for 2 min at 4 °C between washes, and the supernatant decanted. Beads were then resuspended in 100 µL 50 mM NH₄HCO₃ and stored at −20 °C.

The resultant bead mixture was added to 0.5% Rapigest (Waters) in 50 mM NH₄HCO₃ and reduced with dithiothreitol at 65 °C for 30 min. 2-chloroacetamide was then added and the mixture was incubated in the dark for 20 min at room temperature. Mixtures were centrifuged at 200 × g for 2 min at room temperature to pellet beads. The supernatant was decanted and trypsinized with 2.5 µg of trypsin overnight at 37 °C. Mixtures were then concentrated to 100 µL in a speedvac and desalted with C18 desalting columns (Thermo Scientific). Samples were reconcentrated using the speedvac and 100 µL LC-Optima MS grade water was added to solubilize samples. Samples were extracted with ethyl acetate and concentrated in the speedvac. The Pierce QFP assay (Thermo) was used to quantify and normalize peptides.

Trypsinized peptides were separated using reverse-phase nano-high-performance liquid chromatography (nano-HPLC) coupled with a nanoACQUITY ultraperformance liquid chromatography (UPLC) system (Waters Corporation). Peptides were trapped and separated in a 2 cm column (Pepmap 100; 3-m particle size and 100-Å pore size), and a 25-cm EASYspray analytical column (75-m inside diameter [i.d.], 2.0-m C18 particle size, and 100-Å pore size) at 300 nL/min and 35 °C, respectively. A 60 min. gradient of 2% to 25% buffer B (0.1% formic acid in acetonitrile) was conducted on an Orbitrap Fusion Lumos mass spectrometer (Thermo Scientific) with ion source at 2.4 kV and ion transfer tube at 300 °C. MS scans from 350 to 2000 m/z were acquired using the Orbitrap at a resolution of 120,000 and 1e6 AGC target. MS2 spectra were collected with 1.6 m/z isolation width and were analyzed using the 3 s TopSpeed CHOPIN method by the Orbitrap or the linear ion trap depending on peak charge and intensity⁴⁶. Orbitrap MS2 scans were acquired at 7500 resolution with a 5e4 AGC and 22 ms maximum injection time after HCD fragmentation with normalized energy of 30%. Rapid linear ion trap MS2 scans were obtained with a 4e3 AGC, 250 ms maximum injection time after CID 30 fragmentation. Precursor ions were chosen based on intensity thresholds (>1e3) from the full scan and on charge states with a 30-s dynamic exclusion window. Polysiloxane 371.10124 was used as the lock mass. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE⁴⁷ partner repository with the dataset identifier PXD025887.

Data were processed using Metalab verson 1.1.1⁴⁸ with MaxQuant version 1.6.2.3⁴⁹ to identify peptides and protein groups. The integrated reference catalog of the human gut microbiome database⁵⁰ combined with the UniProtKB/Swiss-Prot human sequence database (downloaded Feb 1, 2017)⁵¹ with total 9,920,788 sequences was used as the database search. Search parameters were static carbamidomethyl cysteine modification, specific trypsin digestion with up to two missed cleavages, variable protein N-terminal acetylation and methionine oxidation, match between runs, and label-free quantification (LFQ) with a minimum ratio count of 2. A false discovery rate (FDR) of 1% was used for filtering protein identifications, and potential contaminants and decoys were removed.

GUS enzymes in each database were identified by pairwise alignment to the representative GUS proteins Escherichia coli (EcGUS, UniProt: P05804), Clostridium perfringens (CpGUS, UniProt: Q8VNV4), Streptococcus agalactiae (SaGUS, UniProt: Q8E0N2), and Bacteroides fragilis (BfGUS, PDB: 3CMG). A sequence identity threshold 28% was required with at least one of the four representative proteins. In addition, all conserved residues had to be present and correctly aligned to the representative protein that passed the identity threshold. The conserved residues were: EcGUS E413, E504, N566, K568; CpGUS E412, E505, N567, K569; SaGUS E408, E501, N563, K565; and BfGUS E395, E476, N547, K549. GUS loop classes were determined by multiple sequence alignment with representative proteins, followed by examination of each sequence for specific loop criteria as defined by Pollet et al.²⁷ and Pellock et al.⁵².

In vitro UDH assay

TCS-G was resuspended in 100% DMSO to a concentration of 10 mM. The assay reaction mixture consisted of 10 µL NAD+ (2 mM final), 5 µL uronate dehydrogenase (1 µM final), 5 µL various GUSs (50 nM final), and 30 µL TCS-G (200 µM final). Components were previously diluted in assay buffer (50 mM HEPES, 50 mM NaCl, various pH) or (50 mM sodium acetate, 50 mM NaCl, various pH). The pH of each reaction was determined using the optimal pH of the reaction as determined using pNPG⁵³. Reactions were incubated at 37 °C for 30 min, and absorbance was monitored continuously at 340 nm using a BMG Labtech PHERAstar plate reader. The initial velocity of the reaction was fit using linear regression in MATLAB. Rates are the average of three biological replicates ±SEM.

Catalytic efficiency assay

Assay mixtures contained 10 μL GUS (various final concentrations, between 10–50 nM), 30 μL TCS-G (final concentrations between 30–120 μM), and 10 μL assay buffer (50 mM HEPES, 50 mM NaCl, various pH) or (50 mM sodium acetate, 50 mM NaCl, various pH). Control reactions replaced GUS with buffer. Reactions were quenched at five time points with 50 μL 25% trichloroacetic acid. Samples were centrifuged for 10 min at 16,000 × g, and the supernatant was subjected to analysis by HPLC on an Agilent 1260 Infinity II system using an Agilent InfinityLab Poroshell 120 C18 column (4.6 × 100 mm, 0.7 μm particle size). The column temperature was set to 38 °C with a flow rate of 0.9 ml/min and injection volume of 40 μL. Conditions were set to flow 98% A (water with 0.1% formic acid) and 2% B (acetonitrile with 0.1% formic acid) for two minutes. A linear gradient was then set to flow to 98% B over 10 min and held for 4 min. Conditions were then ramped down to 98% A for 1 min and re-equilibrated at 98% A for 2 min. Analytes were detected using an Agilent DAD detector at a wavelength of 280 nm. Concentrations of TCS-G were determined using a standard curve of TCS-G (0-250 μM). Reaction curves were fit using linear regression, and the resultant initial velocities were plotted against substrate concentration to determine k_cat/K_M. Reported catalytic efficiencies are the average of three biological replicates ±SEM.

In vitro IC₅₀ assay

Reaction mixtures containing 10 µL GUS (10 nM final), 10 µL TCS-G (200 µM final), 5 µL inhibitor (various concentrations), and 25 µL buffer (50 mM HEPES, 50 mM NaCl, various pH) or (50 mM sodium acetate, 50 mM NaCl, various pH) were incubated for 10 min and quenched with 50 µL 25% trichloroacetic acid. Samples were centrifuged 10 min at 16,000 × g, and the supernatant was analyzed using the method described for the catalytic efficiency assay. Inhibition was calculated by the equation below:

where AUC_min is the signal of the uninhibited reaction, AUC_max is the signal of the 100% inhibited reaction, and AUC_inh is the signal of the reaction at a given concentration of inhibitor. Percent inhibition values were plotted against the log of inhibitor concentration, and GraphPad Prism 8.0 was used to determine IC₅₀ values.

In fimo assay

Reaction mixtures contained 5 µL fecal extract (0.1 mg/mL final), 30 µL TCS-G (200 µM final), and 15 µL assay buffer (25 mM HEPES, 25 mM NaCl, pH 6.5). Reactions were quenched at five time points with 50 µL 25% trichloroacetic acid. Samples were centrifuged 20 min at 16,000 × g in a tabletop centrifuge, and the supernatant was analyzed via the same HPLC method described for the catalytic efficiency assay. Reaction rates were determined by fitting progress curves using linear regression and are expressed as initial turnover rates (µM/s). Controls contained fecal extract that had been heat-killed at 95 °C for 5 min. Reported rates are the average of three biological replicates ±SEM.

In fimo inhibition assays

Reaction mixtures contained 5 µL fecal extract (0.1 mg/mL final), 10 µL TCS-G (200 µM final), 5 µL GUSi (10, 1, or 0.1 µM final), and 30 µL assay buffer (25 mM HEPES, 25 mM NaCl, pH 6.5). Reactions were quenched at 30 min with 50 µL 25% trichloroacetic acid. Samples were centrifuged 20 min at 16,000 × g in a tabletop centrifuge, and the supernatant was analyzed via the same HPLC method described for the catalytic efficiency assay. Controls contained 5 µL buffer in place of GUSi. Inhibition was calculated using Eq. 1.

Crystallography

Crystals were produced using the sitting drop vapor diffusion method at 20 °C. Trays were set up using the Art Robbins Instruments Crystal Phoenix robot or an Oryx4 robot (Douglas Instruments) and Hampton Research three-well midi crystallization plates (Swissci). For Roseburia hominis 3 GUS, crystals were produced in a condition containing 100 nL 12.5 mg/mL Rh3 GUS and 200 nL 0.2 M LiCl, 20% PEG 3350. For Faecalibacterium prausnizii L2-1 GUS, FpL2-1 GUS at 15 mg/mL was preincubated with UNC10201652 (GUSi) and PNPG in tenfold excess prior to addition to the crystalline solution. Crystals formed in a condition containing 200 nL GUS and 100 nL 0.2 M potassium thiocyanate (KSCN), 20%(w/v) PEG 3350.

Crystals were cryo-protected using the crystal solutions as described above with 20% glycerol. Diffraction data were collected at 100 K at APS beamline 23-ID-D. Data were processed using XDS and structures were solved using molecular replacement in Phenix. For Rh3 GUS, the R. hominis 2 structure (PDB 6MVH) was used as a search model. For FpL2-1 GUS, a FpL2-1 model produced using the Phyre2 server was used as a search model⁵⁴. Maps and models output from molecular replacement were run through the Autobuild function of Phenix (version 1.17.1-3660)⁵⁵. Structures were refined using phenix.refine, and Coot (version 0.9.4) was used for manual, visual inspection, and ligand fitting⁵⁶. Final PDB coordinates were deposited to the RCSB Protein Data Bank under the codes 7KGZ (Rh3) and 7KGY (FpL2-1 with GUSi).

Docking of TCS-G using Schrodinger

TCS-G docking into various GUS enzymes was carried out using the Schrödinger (Release 2020-1, http://www.schrodinger.com) induced fit docking pipeline. The Schrödinger Protein Preparation module was used to prepare proteins for docking by adding hydrogens, deleting water molecules more than 3 Å from the ligand, generating protonation states based on the protein’s ideal pH (as determined previously via in vitro assays), and creating metal and disulfide bonds. Default settings were used with the exception of the pH protonation states. The wizard was used to preprocess the structure, remove waters, optimize H-bonds, and minimize the structure.

Ligands were prepared using the Ligprep module. Ionization states were generated at pH 6.5 ± 0.5. Induced fit docking was used to dock TCS-G into the active sites of each GUS enzyme. Using a previously solved structure of E. eligens GUS with glucuronic acid bound (PDB: 6BJQ), glucuronic acid was added into the active site of each GUS using PyMOL by aligning each GUS to 6BJQ. The box center was then chosen as the glucuronic acid location with a box size of 30 Å. Core constraints were added to restrict docking of TCS-G to the existing glucuronic acid structure using the maximum common substructure. Glide redocking was performed at XP precision. Top docking poses were chosen based on the IFD Docking score and the Glide Score, as well as visual examination to confirm probable binding mode.

Circular dichroism

The protein stabilities of Rh3 GUS and Fp2-L1 GUS and their mutants were determined using the circular dichroism method⁵⁷. Enzyme (0.125 mg/mL) in CD buffer (10 mM potassium phosphate pH 7.4, 100 mM potassium fluoride) was loaded into a 1-mm cuvette. The Chirascan Plus instrument (Applied Photophysis Limited) was used to acquire 1) scan spectra from 185 to 260 nM at 20 °C and 2) a melting profile at 193 nm from 20 to 94 °C. Spectra acquired with buffer alone were used to correct for background signal.

Cell culture and treatment with TCS and TCS-G

MC38 intestinal epithelial cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum (FBS) and incubated at 37 °C under an atmosphere of 5% CO₂. MC38 cells were seeded at 20% confluency and left to settle overnight. Cells were then treated with either 1 μM TCS, TCS-G, or vehicle (DMSO). After 48 h, cells were harvested for RT-qPCR analysis. Cell medium was collected for ELISA analysis using the CBA Mouse Inflammation Kit (BD Sciences) according to the manufacturer’s instructions, and data were acquired using a BD LSR Fortessa flow cytometer (BD Biosciences) and analyzed using FlowJo software (Treestar).

ELISA of inflammatory biomarkers in plasma

Blood samples were harvested via cardiac puncture and collected in blood collection tubes (Covidien). The plasma fractions were prepared by centrifugation of the harvested blood at 3000 × g for 5 min at 4 °C. The concentrations of cytokines in plasma were determined using the CBA Mouse Inflammation Kit (BD Biosciences) as described above.

Reverse-transcriptase-qPCR of inflammatory biomarkers

Total RNA of colon tissues and MC38 cells were isolated using Trizol reagent (Ambion) according to the manufacturer’s instruction. RNA was reverse transcribed into cDNA using the High Capacity cDNA Reverse Transcription kit (Applied Biosystem) according to the manufacturer’s instructions. In all, 20 μL PCR reactions were prepared using the Maxima SYBR Green Master Mix (Thermo Fisher Scientific), and qPCR was carried out using a DNA Engine Opticon System (Bio-Rad Laboratories). Mouse-specific primer sequences (Thermo Fisher Scientific) used to detect inflammatory biomarkers are listed in Supplementary Table S6. Gapdh expression was used as an internal control.

Flow cytometry

Distal colon tissues were dissected, washed with cold PBS, and digested with Hank’s balanced salt solution (HBSS, Lonza) supplemented with 1 mM dithiothreitol (DTT) and 5 mM EDTA at 4 °C (colon epidermal cells). The released cells were stained with FITC-conjugated anti-mouse CD45 (BioLegend, Clone: 30-F11), PerCP/Cy5.5-conjugated anti-mouse F4/80 (BioLegend, Clone: BM8), and PE/Cy7-conjugated anti-mouse Ly-6G/Ly-6C (GR-1) (BioLegend, Clone: RB6-8C5) with a 1:100 diluted solution. Cells were stained with Zombie Violet^TM dye (Zombie Violet^TM Fixable Viability Kit; BioLegend) according to the manufacturer’s instructions to exclude dead cells. Gating and cell identification strategies are as follows: briefly, cell doublets and clumps were eliminated using FSC-A gating and debris was eliminated using FSC-A vs SSC-A. Dead cells were gated out using Zombie Violet™ dye. Flow cytometry data were acquired on a BD LSR Fortessa^TMcell analyzer (Becton Dickinson, Franklin Lakes, NJ) and analyzed using FlowJo software (FlowJo, LLC). Gating strategies used for the identification of major immune cell populations are shown in Supplementary Fig. S17.

Histological staining

The dissected colon tissues were fixed in 10% neutral buffered formalin (Thermo Fisher Scientific) for 48 h. After dehydration, the tissues were embedded in paraffin and sliced (5 mm) by Rotary Microtome (Thermo Fisher Scientific). The slices were dewaxed in serial xylene and rehydrated through ethanol solutions, stained with hematoxylin and eosin (Sigma-Aldrich), and images were obtained under 200× magnification (BZ-X700 microscope, Keyence, Itasca, IL). The histologic scores were evaluated by a blinded observer according to the following measures: crypt architecture, degree of inflammatory cell infiltration, muscle thickening, goblet cell depletion, and crypt abscess. The histologic damage score is the sum of each individual score.

DNA extraction

DNA was extracted from mouse fecal samples using QIAmp DNA Stool Mini Kit (Qiagen, Valencia, CA) following instructions from the manufacturer with an additional bead-beating step. The quantity of the extracted DNA was measured using a NanoDrop Spectrophotometer (Thermo Fisher Scientific), and the quality was verified using gel electrophoresis. The DNA was then subjected to further analysis.

Real-time PCR analysis of 16S rRNA gene

DNA extracted from mouse fecal samples were subjected to qPCR analysis using a DNA Engine Opticon System (Bio-Rad Laboratories, Hercules, CA). In all, 20 μL PCR reactions were made using the Maxima SYBR_green Master Mix (Thermo Fisher Scientific), and DNA was normalized to 5 ng/μL per reaction. The 16S rRNA primers are in Supplementary Table S6.

16S rRNA sequencing and analysis

DNA quality was monitored on 1% agarose gels. The V3–V4 hypervariable regions of the bacteria 16S rRNA gene were amplified with primers 341F (5’-CCTAYGGGRBGCASCAG-3’) and 806R (5’-GGACTACNNGGGTATCTAAT-3’). PCR products were detected on 2% agarose gels by electrophoresis and purified using the Qiagen Gel Extraction Kit (Qiagen, Germany). Sequencing libraries were generated using NEBNext Ultra DNA Library Pre Kit for Illumina, following the manufacturer’s recommendations and index codes were added. The library quality was assessed using the Qubit 2.0 Fluorometer (Thermo Scientific) and Agilent Bioanalyzer 2100 system. The library was sequenced on an Illumina platform and 250 bp paired-end reads were generated.

To analysis difference in abundance patterns among samples, beta diversity using weighted UniFrac distance followed by Principal Coordinate Analysis (PCoA) and non-metric multidimensional scaling (NMDS) analysis. The alpha diversity was considered as the richness of the samples, number of OTUs present per treatment. We calculated Simpson, Shannon, Chao1 index using QIIME2 version 2019.7.0⁵⁸ and Phytools package 0.7 v⁵⁹ in R (R Development Core Team, 2014). Statistical tests were performed in R. Kruskal-Wallis rank-sum test was used to compare alpha diversity indexes among treatments. The beta diversity was calculated using Phytools 0.7 v. The weighted UniFrac distance method was used to create the matrix followed by PCoA to compare similarity among treatments. All plots were obtained using ggplot2⁶⁰.

Cell proliferation assay

Mouse (MC38) or human (Caco2 and HCT-116) intestinal cells were grown in DMEM medium fortified with 10% FBS (EMD Millipore Corporation). The cells were seeded in 96-well plates, then treated with GUSi or vehicle (0.2% v/v DMSO) for 24 h. Cell viability was determined using an MTT assay.

Data and statistical analyses

Data are mean ± SEM. For the comparison between two groups, Shapiro–Wilk test was used to verify the normality of data; when data were normally distributed, statistical significance was determined using two-sided t test; otherwise, significance was determined by Wilcoxon–Mann–Whitney test. The statistical comparison of three groups was analyzed using one-way ANOVA by Tukey’s multiple comparisons. The statistical analyses were performed using SAS (version 9.3) statistical software and GraphPad Prism (version 8.0 or 9.0). P < 0.05 was considered statistically significant.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.