Target prediction and validation

TGR5 (Human GPBAR1-(AA 1-330)-Q8TDU6-Full length, MW 36800, cat. no. CSB-CF819471HU) was purchased from Yangene Biotechnology Co., Ltd. We searched for SMILES of butyric acid (MW 88.11, cat. no. B103500, Sigma-Aldrich, Merck KGaA) through PubChem and obtained potential targets for butyric acid by SwissTargetPrediction database inputting SMILES of butyric acid and selecting Homo sapiens. We validated the target of butyric acid through surface plasmon resonance (SPR, Biacore T200, GE) assay, as follows: The activator was prepared by mixing 400 mM 1-Ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC, cat. no. E1769-25G, Sigma-Aldrich, Merck KGaA) and 100 mM N-Hydroxysuccinimide (NHS, cat. no. 130672-25G, Sigma-Aldrich, Merck KGaA) immediately prior to injection. The CM5 sensor chip (cat. no. 29149603, GE) was activated for 600 s with the mixture at a flow rate of 10 µL/min. We used immobilization buffer (10 mM Sodium Acetate, pH 5.0) to dilute the target protein to 50 µg/mL and injected it to sample channel (Fc4) at a flow rate of 10 µL/min to reach immobilization levels of 2400 resonance units (RU). The chip was deactivated by 1 M Ethanolamine hydrochloride (cat. no. E6133-100G, Sigma-Aldrich, Merck KGaA) at a flow rate of 10 µL/min for 420 s. We utilized the analyte buffer [HEPES (pH 7.4): 10 mM HEPES, 150 mM NaCl, 3 mM EDTA, 0.05% P20] to dilute butyric acid (cat. no. B103500, Sigma-Aldrich, Merck KGaA) to different concentrations (125, 62.5, 31.25, 15.625, 7.8, and 0 µM). Butyric acid was injected to channel Fc1- Fc4 at a flow rate of 30 µL/min for an association phase of 90 s, followed by 180 s dissociation. The association and dissociation process were carried out in the analyte buffer. We repeat 6 cycles in the order of increasing analyte concentration.

Limited proteolysis-mass spectrometry

Sample preparation: The experiment is divided into two groups: butyric acid group and control group. We mixed butyric acid (100 µM) or equal volume solvent with TGR5 (10 µg) and incubated them for 15 min at 25℃. Protease K was added to protein solution at a proteinase K to protein mass ratio of 1:100. We incubated the mixture for 3 min at 25℃, and immediately heat it at 98℃ for 5 min to terminate the digestion reactions. The samples after enzymatic hydrolysis were cooled to room temperature, and then equal volume of 2% sodium deoxycholate (SDC, dissolved by 20 mM Tris HCl) was added. The pH was adjusted to 7-8.5 with ammonium bicarbonate. The samples were heated at 98℃ for 5 min, cooled to room temperature and then reacted with 5 µL TCEP (0.1 M) and 5 µL chloroacetamide (0.4 M) in a dark environment at 45℃ and 1500 rpm for 5 min. The samples were taken out and cooled to room temperature, then were digested by trypsin (Promega, Madison, WI) at an enzyme substrate ratio of 1:50 overnight at 37℃. The samples were added an appropriate amount of formic acid (final concentration of 1.5%), mixed well and centrifuged at 16,000 g for 5 min to obtain the supernatant. Then the mixtures were desalted by C18 desalination column, vacuum dried, solubilized in 0.1% formic acid, and immediately analyzed by mass spectrometry using Exploris 480 mass spectrometer (Thermo Fisher Scientific, MA, USA). Tandem mass spectrometry was analyzed by Proteome Discoverer (Thermo Fisher Scientific, MA, USA).

Molecular docking

The AlphaFold Protein Structure Database was used to obtain the three-dimensional structure of TGR5 through the input of UniProt accession (Q8TDU6). Molecular docking was completed by Covalent Docking in the Glide module of Schrödinger Maestro software. The Protein Preparation Wizard module was used to prepare, optimize, and minimize the receptor through constrained minimization using the OPLS3e force field. All molecules were prepared according to the default settings of the LigPrep module. When performing screening in the Glide module, the prepared receptor was imported to specify the appropriate position for receptor grid generation. The predicted site of the protein was selected as the centroid of a 12 Å box. Finally, molecular docking and screening were performed through standard docking methods.

Molecular dynamics (MD) simulation

Gromacs 2020 software package was employed to perform molecular dynamics simulation for the screened receptor-ligand complex. The protein employed the AMBER99SB-ILDN force field parameters, while the small-molecule ligand adopted the GAFF force field parameters, and the topology of the small molecule was constructed using the Sobtop program, along with charge fitting conducted by RESP. The TIP3P water model was selected, with the minimum distance between the protein and the edge of the water box being 1.0 nm. The system charge was neutralized using sodium ions or chloride ions according to the docking results. The molecular dynamics simulation workflow consists of four steps, namely, energy minimization, heating, equilibration, and production dynamics simulation. The calculation of binding free energy between the ligand and the protein employed the gmx_MMPBSA method within the Gromacs 2020 program.

Cells and preparation of medium

The RAW264.7 cells were purchased from Procell Life Science&Technology Co., Ltd (Wuhan, Hubei province, China) and were cultured in DMEM medium (Gibco) supplemented with 10% fetal bovine serum (FBS, Gibco), 1% Penicillin-Streptomycin Solution at 37 °C in a humidified 5% CO2 atmosphere. LPS (100 ng/mL, Sigma-Aldrich; Merck KGaA) was used to induced siRNA-β-arrestin2-transfected and non-transfected RAW264.7 cell polarization for 24 h. SB (50 µg/mL) was administered to the RAW264.7 cells followed by 2 h after LPS administration. INT-777 (cat. no. HY-15677, MCE, USA), a TGR5 agonist, was mixed with DMSO and diluted to a final concentration of 3 µM, followed by 2 h after LPS administration. SBI-115 (cat. no. HY-111534, MCE, USA), a TGR5 antagonist, was mixed with DMSO and diluted to a final concentration of 100 µM, followed by 2 h after LPS administration.

Cells transfection with siRNA

siRNA-β-arrestin2 was transfected into RAW264.7 cells to knock down β-arrestin2 in RAW264.7 cells. siRNA-β-arrestin2 and negative siRNA-control (siRNA-NC) were synthesized by Jintuosi Biotechnology Co., Ltd in Wuhan, China. Three siRNA sequences were designed for the β-arrestin2 target gene as follows: siRNA-Arrb2(M)-568: sense: 5’-GGCUUGUGGAGUAGACUUUTT-3’ antisense: 5’-AAAGUCUACUCCACAAGCCTT-3’; siRNA-Arrb2(M)-916: sense: 5’-GGCUCAGCUAGAACAAGAUTT-3’ antisense: 5’-AUCUUGUUCUAGCUGAGCCTT-3’; siRNA-Arrb2(M)-1024: sense: 5’-GGAUGGGCAGCUCAAACAUTT-3’ antisense: 5’-AUGUUUGAGCUGCCCAUCCTT-3’. RAW264.7 cells were incubated in 6 well plates at 37 °C in 5% CO2 atmosphere overnight and cultured in serum-free DMEM medium 2 h before transfection. We used 100 µL serum-free opti-MEM (31985-070, Gibco) to dilute 10 µL siRNA (20 µM) and 5 µL LipofectamineTM 2000 (11668-019, Invitrogen) respectively and allowed the solution to stand at room temperature for 5 min. We mixed a diluent of LipofectamineTM 2000 and siRNA and allowed the solution to stand at room temperature for 20 min. RAW264.7 cells were incubated with siRNA diluent at 37 °C in 5% CO2 atmosphere for 6 h and then cultured in DMEM medium with FBS. After 24 h of transfection, the siRNA knockdown efficiency was identified using qRT-PCR.

Quantitative real-time PCR (qRT-PCR)

We used qRT-PCR to detect mRNA expression of macrophage polarization markers. We used Trizol reagent (cat. no. 15596026; Invitrogen; Thermo Fisher Scientific, Inc.) to extract total RNA from RAW264.7 cells. We utilized HiScript II Q Select RT SuperMix for qPCR kit (cat. no. R233; Vazyme Biotech Co., Ltd.) to get rid of DNA for 2 min at 42 ˚C and convert RNA to cDNA according to following conditions: 50˚C for 15 min, 85˚C for 5 s, 4˚C for 10 min. Then we used AceQ qPCR SYBR Green Master Mix kit (cat. no. Q111‑02; Vazyme Biotech Co., Ltd.) for qRT-PCR detection on a QuantStudio 6 Flex Real‑Time PCR system (Applied Biosystems; Thermo Fisher Scientific, Inc.) according to following conditions: 95˚C for 10 min, 40 cycles of 95˚C for 15 s and 60˚C for 30 s, 95˚C for 15 s and 60˚C for 60 s and 95˚C for 15 s. We adopt 2-ΔΔ Ct method for quantitative analysis of the data. The primer sequences for target genes were displayed in Table 1.

Immunofluorescence staining

We rinsed the round coverslips with cultured cells using phosphate buffered saline (PBS) in the plate. The cultured cells were placed in 4% paraformaldehyde (cat.no.80096618, Sinopharm Chemical Reagent Co., Ltd.) for 15 min and rinsed for 3 times with PBS. Next, the fixed cells were permeabilized with 0.5% Triton X-100 (cat.no.ST795, Beyotime, China) for 20 min at room temperature, and rinsed for 3 times with PBS. We added 10% normal Goat Serum (cat.no.AR1009, Wuhan Boster Biological Technology co., Ltd) on the round coverslips for 30 min at room temperature. Then, cells were added enough diluted primary antibody: CD86 (cat.no.13395-1-AP, Wuhan Proteintech Group, Inc.), CD206 (cat.no.18704-1-AP, Wuhan Proteintech Group, Inc.) at 4 °C overnight, rinsed with PBS, and incubated with CY3 Conjugated AffiniPure Goat Anti-Rabbit IgG (H + L) (BA1032, Wuhan Boster Biological Technology co., Ltd) for 1 h at 37 °C. Cells were incubated with DAPI (cat.no.C1002, Beyotime, China) for 5 min in dark. The images were captured with a fluorescence microscope (Olympus, Yokohama, Japan).

Western blot analysis

After RAW264.7 cells were administered with LPS with or without INT-777 and SBI-115 for 24 h, we utilized phenylmethanesulfonyl fluoride (PMSF, P105539, Shanghai Aladdin Bio‑Chem Technology Co., Ltd) to lyse cells and extract the total protein. BCA kit (P0010, Beyotime, China) was used to detect the protein concentration. The protein was separated by electrophoresis through 12% SDS-PAGE gel and transferred to polyvinylidene difluoride (PVDF) membranes. Then PVDF membranes were incubated with Tris buffered saline containing 0.1% Tween‑20 (TBST) including 5% skimmed milk for 2 h at room temperature. Primary antibodies were used to incubate the membranes overnight at 4 °C. Primary antibodies were as follows: GAPDH (37 kDa, AB-P-R 001, Hangzhou Goodhere Biotechnology Co., Ltd.), TGR5 (33 kDa, ab72608, Abcam), β-arrestin2 (50 kDa, ab54790, Abcam). The PVDF membranes were washed by TBST and further incubated with HRP-conjugated AffiniPure Goat anti-mouse secondary antibody (BA1051; Wuhan Boster Biological Technology, Ltd.) and HRP conjugated AffiniPure goat antirabbit secondary antibody (BA1054; Wuhan Boster Biological Technology, Ltd.) for 2 h at room temperature. Finally, ECL kit (P1050; Beijing Applygen Technologies Inc.) was used to visualize the protein bands. We made use of BandScan v5.0 software (Glyko Biomedical Ltd.) to analyse gray-scale values of the blots.

Statistical analysis

All data were presented as mean ± standard deviation. Student’s t test was used to analyse the differences between two groups. Differences among multiple groups were analysed by one‑way ANOVA, followed by Tukey’s post‑hoc test. P-value < 0.05 was statistically significant. Part figures were made in STRING (version 12.0) (string-db.org).