Activation of Free Fatty Acid Receptor 4 Affects Intestinal Inflammation and Improves Colon Permeability in Mice

The RAW264.7 murine macrophages (ATCC: TIB-71) were cultured in Dulbecco’s Modified Eagle Medium (Gibco) supplemented with 10% bovine calf serum (BCS), 2 mM Alanine-Glutamine, 0.5% penicillin-streptomycin (P/S), 1 mM sodium pyruvate, and 25 mM HEPES. Cells were grown in an atmosphere of 5% CO₂ at 37 °C. The cells were passaged after reaching approx. 80% confluence.

The Caco-2 human colonic epithelial cell line was obtained from the American Type Culture Collection (ATCC: HTB-37). The cells were cultured in Eagle’s Minimal Essential Medium (Gibco) with addition of 20% fetal bovine serum (FBS), 4 mM Ala-Gln, 0.1 mM nonessential amino acids, 0.5% P/S, 1 mM sodium pyruvate, and 25 mM HEPES. Cells were grown in an atmosphere of 5% CO₂ at 37 °C. The cells were trypsinized and passaged after reaching approx. 80% confluence.

The toxicity in RAW264.7 cells was evaluated with neutral-red uptake (NRU) test, which characterizes the lysosomal activity of viable cells [18]. The cells were placed in 96-well plates (20,000 cells/well). FFAR agonists in the concentrations of 10, 20, 50, and 80 µM were chosen based on the preliminary experiments (data not shown), and budesonide was applied at the concentration of 10 µM. Cells were exposed to the compounds for 48 h before the measurement of cytotoxicity. Then, cells were incubated for 1 h with 100 μL/well of 0.05 mg/mL NR solution in culture medium. Next, cells were washed with phosphate-buffer saline (PBS, pH 7.4) and 100 μL/well of 40% ethanol, and 10% acetic acid in water was used to dissolve the dye. After 10 min of shaking, the absorbance was measured at 540 nm in a microplate reader (iMARK Microplate Reader, Biorad, Hertfordshire, UK). Data were expressed as a percentage of cells without any treatment.

The cytotoxicity in Caco-2 cell line was evaluated with the MTT assay [19]. Briefly, cells were seeded on a 96-well plate at the density of 20,000 cells/well and incubated for 24 h with FFAR agonists in the following concentrations: 10, 30, 50, 100, and 200 µM. Budesonide was applied at the concentration of 10 µM. Then, 20 µL of MTT solution (5 mg/mL) was added to each well and incubated for 2 h. Next, medium was replaced with 100 µL of dimethyl sulfoxide (DMSO), and the plate was shaken for 15 min. Optical density was measured by microplate reader (iMARK Microplate Reader, Biorad, Hertfordshire, UK) at 595 nm. Data were expressed as a percentage of cells without any treatment.

The cells were placed in 96-well plates (20,000 cells/well) and incubated for 24 h with standard culture medium (control) or medium with 1 μg/mL lipopolysaccharide (LPS) with or without FFAR agonists (in the same concentrations as for cytotoxicity assessment) or budesonide. After 24 h of incubation, LPS was removed from the culture by replacing the media with new solutions containing FFAR agonists or budesonide but without LPS. After another 24 h, 100 µL of cell culture supernatant was mixed with 100 µL of Griess reagent water solution (40 mg/mL), and the mixture was kept in the dark for 15 min. Next, the absorbance was read at 540 nm. Nitrite concentration was expressed as a percentage of the cells treated with LPS only.

The model was adapted from the study by Van De Walle et al. [20]. Briefly, the Caco-2 cells were seeded on a 24-well plate at the density of 100,000 cells/well and incubated for 24 h with medium. Then, a mixture of tumor necrosis factor α (TNFα; 50 ng/mL), interleukin 1β (IL-1β; 25 ng/mL), interferon γ (IFNγ; 50 ng/mL), and LPS (100 ng/mL) was added to each well. Experimental groups were additionally treated with FFAR agonists in concentrations selected based on the outcomes of cytotoxicity assessment: FFAR1 100 µM, FFAR2 30 µM, FFAR3 50 µM, and FFAR4 30 µM. After 12 h of incubation, cell culture medium was harvested and used for IL-6 measurement with a human IL-6 ELISA kit (cat. No. 950.030.096, Diaclone, France) according to manufacturer instructions. Outcomes were expressed as a percentage of cells treated with a mixture of cytokines and LPS only.

We used male C57BL/6 mice (Mossakowski Medical Research Center Polish Academy of Sciences in Warsaw, Poland) weighing 22–26 g (6–8 weeks of age). Mice were kept at 22 °C and under a 12-h light/dark cycle in sawdust-lined plastic cages. Chow pellets and tap water were provided ad libitum. All animal protocols were approved by the Medical University of Lodz Animal Care Committee (Protocol 34/ŁB 96/2018) and complied with the European Communities Council Directive of 22 September 2010 the EU (2010/63/EU).

To assess therapeutic activity of FFARs, in both DSS and TNBS model, mice were injected intraperitoneally (i.p) twice daily for 4 consecutive days (on days 3–6). The following agents were used at the dose 1 mg/kg bw: GW9508 (FFAR1 agonist), 4-CMTB (FFAR2 agonist), AR420626↗ (FFAR3 agonist), and GSK137647 (FFAR4 agonist). AH7614 (FFAR4 antagonist) was administered at the dose of 5 mg/kg bw.

MPO activity was measured according to the protocol described earlier [21]. Briefly, 1-cm segments of the colon were weighed, homogenized in hexadecyltrimethylammonium bromide (HTAB) buffer, and centrifuged (13,200× g, 15 min). Then, 7 μL of supernatants were added to each well on a 96-well plate containing 200 μL of 500 mM potassium phosphate buffer supplemented with 0.167 mg/mL of O-dianisidine hydrochloride and 0.05 μL of 1% H₂O₂. Absorbance (450 nm; triplicate) was measured 0, 30, and 60 s after the initiation of reaction. MPO activity was expressed in milliunits per gram of wet tissue, 1 unit being the quantity of enzyme able to convert 1 μmol of H₂O₂ to water in 1 min at room temperature (RT). Standard curve of MPO activity was made with purified peroxidase enzyme.

For immunohistochemical (IHC) staining, the method described by Sadej et al. [23] was used. Briefly, slides were washed in PBS followed, by quenching of endogenous peroxidase activity with PBS containing 0.6% (v/v) H₂O₂ and 10% (v/v) methanol for 30 min at RT. Next, slides were washed and immersed in blocking solution/antibody diluent containing 2% (w/v) bovine serum albumin (Sigma-Aldrich, St. Louis, MO, USA), 5% (v/v) goat serum (Dako, Carpinteria, CA, USA), and 0.2% (v/v) Triton X-100 in PBS for 1 h. The sections were incubated overnight at 4°C with primary antibodies to visualize FFAR1 (AFR-001, 1:200; Alomone Labs, Jerusalem, Israel), FFAR2 (AFR-032; 1:500; Alomone Labs, Jerusalem, Israel), FFAR3 (PA5-97747; dil. 1:200; Thermo Scientific, Rockford, IL, USA), or FFAR4 (NBP1-00858; 1:2000; Bio-Techne, Warsaw, Poland).

The following day, slides were washed extensively for 3 × 10 min in PBS and further incubated with the appropriate secondary antibody for 1.5 h at RT. After another round of washing, slides were incubated with an avidin–biotin complex reagent (1:200; Vector Laboratories, Burlingame, CA, USA) for 1 h at RT. The staining was visualized by incubating the slides with 0.1% (w/v) 3,3-diaminobenzidine and 0.03% (v/v) H₂O₂ in Tris-HCl, pH 7.6. The reaction was terminated after approx. 4–5 min. Then slides were dehydrated through a graded series of ethanol, cleared in xylene, and cover-slipped in Pertex mounting medium (HistoLab, Göteborg, Sweden). Evaluation of the staining was performed in the computerized Axo Imager Carl Zeiss light microscope. To quantify the images, we utilized Fiji software. Spots of FFAR 1–4 staining were separated on the basis of the hue, the saturation, and the brightness. Pixel coverage of the FFARs staining was calculated for epithelium areas. We applied the same cut-off levels in all the analyzed pictures selected to quantify FFARs in our experiment. The amount of FFARs was presented as a surface-area coverage (%).

Epithelial ion transport was evaluated according to the method described previously [24]. Briefly, sections (around 1 cm) of the distal large intestine were obtained and immediately positioned in Ussing chamber (Physiologic Instruments, Inc., San Diego, CA, USA) with 6 mL of Krebs solution of the following constitution (mM): NaCl, 115; KH₂ PO₄, 2; MgCl₂, 2.4; NaHCO₃, 25; KCl, 8; and CaCl₂, 1.3. The solution was saturated with 95% O₂ and 5% CO₂ and supplemented with either glucose (10 mM) or mannitol (10 mM) by the basolateral and mucosal side, respectively. The bath temperature was kept at 37 °C. The exposed area of the tissue was 0.3 cm². Tissues were voltage clamped to zero, using the WPI EVC-4000 voltage clamp apparatus (World Precision Instruments, Sarasota, FL, USA) with Ag/AgCl electrode and 3 M KCl agar bridge. Once a stable baseline in short circuit current (Isc, mA/cm²) was reached (15–30 min), we tested FFAR agonist: GW9508 (10⁻⁵ M, FFAR1 agonist), 4-CMTB (10⁻⁵ M, FFAR2 agonist), AR 420626 (10⁻⁵ M, FFAR3 agonist), and GSK 137647 (10⁻⁵ M, FFAR4 agonist), dissolved in DMSO or an equivalent volume of vehicle (DMSO, final concentration: 0.1%) pipetted to the basolateral side of the chamber. After 10 min, specimens were challenged with either forskolin (FSK) (10⁻⁵ M, cAMP-dependent secretagogue) or veratridine (VER) (3 × 10⁻⁵ M, voltage-dependent Na⁺ channel activator). For each challenge, the peak change in Isc (∆Isc) was established.

For measurements of intestinal permeability, sections of the mouse large intestine were mounted in Ussing chambers (0.3 cm² opening), and mucosal and serosal sides were exposed to 6 mL of oxygenated Krebs buffer. Mannitol and glucose (both 10 mmol) were introduced, respectively, and the tissues were allowed to incubate for 15 min. The experiment was started by replacing the buffer in the mucosal compartment with 10 mL Krebs buffer with fluorescein isothiocyanate dextran (FITC-dextran; Sigma-Aldrich) at a concentration of 1 mg/mL. Then, compartments were protected from light by aluminum foil. Fluorescence in the serosal chamber was measured instantly to establish baseline fluorescence. Specimens were taken from the serosal chambers at 20-min intervals for a total period of 2 h, and FITC-dextran concentration was measured at an excitation wavelength of 494 nm and emission wave-length of 521 nm.

In brief, RNA was sequestered from the distal sections of large intestine (weighing 20–30 mg) from both healthy and DSS-treated animals, in accordance with the manufacturer’s protocol using Total RNA Mini Plus kit (A&A Biotechnology, Gdansk, Poland). RNA was eluted from ion exchange columns by diethyl pyrocarbonate (DEPC)-treated water (40 μL). Then, total RNA was purified using Lithium Chloride Precipitation Solution in accordance with the manufacturer’s protocol (Life Technologies, Carlsbad, CA, USA). The purity and amount of isolated RNA was estimated using Colibri Microvolume Spectrophotometer (Biocompare, San Francisco, CA, USA). Total RNA (2 μg) was transcribed to cDNA with high-capacity Reverse Transcriptase Kit (Life Technologies, Carlsbad, CA, USA) in accordance with the manufacturer’s protocol. Quantitative assay of the expression was executed using fluorescently labeled probes (Life Technologies, Carlsbad, CA, USA): OCLN (Mm00500912_m1), CLDN1 (Mm00516701_m1), CLDN2 (Mm00516703_s1), CLDN3 (Mm00515499_s1), CLDN4 (Mm00515514_s1), CLDN7 (Mm00516817_m1), CLDN10 (Mm01226326_g1), CLDN12 (Mm01316510_m1), and GAPDH (Mm99999915_g1) as endogenous control on Mastercycler S realplex 4 (Eppendorf, Hamburg, Germany) using TaqMan Gene Expression Master Mix (Life Technologies, Carlsbad, CA, USA) according to the manufacturer’s protocol. All experiments were conducted in triplicate. The threshold cycle (Ct) values for studied genes were normalized to Ct values received for GADPH. The relative quantity of mRNA copies was calculated using the equation: 2^−ΔCt × 1000.

Tissue samples and cell culture samples were incubated with the mammalian cell lysis buffer (50 mM Tris−HCl, pH 7.5; 1 mM EDTA, 150 mM NaCl; 0.1% SDS; 0.5% deoxycholic acid; and 1% Igepal CA-630; Sigma-Aldrich, Poznan, Poland) containing protease-inhibitor cocktail. Afterwards, samples were homogenized using Precellys Evolution Homogenizer (Bertin Instruments, Paris, France) and centrifuged at 15,000 g for 15 min at 4 °C to remove the debris. Total protein concentration was measured using the Pierce 660 nm protein assay (Thermo Scientific, Rockford, IL, USA). Electrophoresis was performed using precast 4%–20% SDS-PAGE gel (Bio-Rad, Warsaw, Poland) in the buffer containing 0.1% SDS, 192 mM glycine, 25 mM Tris, pH 8.3. Separated proteins were transferred onto PVDF membranes (pore size, 0.45 μm; Life Technologies, Carlsbad, CA, USA) using a semi-dry system with transfer buffer containing 20% (v/v) methanol, 192 mM glycine, and 25 mM Tris, pH 8.3. The PVDF membranes were then blocked at RT for 1 h in 5% non-fat dry milk in phosphate buffered saline (PBS) with Tween 20 (PBST; PBS, 0.1% Tween 20). Then, membranes were incubated for 80 min at RT with specific primary antibodies diluted in 1% non-fat dry milk in PBST for immunodetection of the proteins of interest. The primary rabbit anti-mouse FFAR1 polyclonal antibody (AFR-001; dil. 1:1000; Alomone Labs, Jerusalem, Israel), FFAR2 polyclonal antibody (AFR-032, dil. 1:1000; Alomone Labs, Jerusalem, Israel), FFAR3 polyclonal antibody (PA5-97747; dil. 1:1000; Thermo Scientific, Rockford, IL, USA), FFAR4 polyclonal antibody (E-AB-31576; dil. 1:1000; Elabscience, Wuhan, China), and mouse anti-glycerylaldehyde-3-phosphate dehydrogenase (β-actin; sc-47778; dil. 1:1000; Santa Cruz Biotechnology, Dallas, TX, USA) were used. After the washing using PBST (5 times, 3 min), membranes were incubated with appropriate secondary antibodies for 1 hat RT. Then bands were visualized using Super Signal west pico western blotting substrate (Thermo Scientific, Rockford, IL, USA) as a substrate for the localization of HRP activity. Qualitative and quantitative analysis was performed by measuring integrated optical density (IOD) by ImageLab v5.2.1 for WindowsTM program (Bio-Rad SA, Warsaw, Poland).

All drugs and reagents, unless otherwise stated, were purchased from Sigma-Aldrich (Poznan, Poland). FFAR ligands were purchased from Bio-Techne (Warsaw, Poland).

Statistical analyses were performed using Prism v9.0 (GraphPad Software Inc., La Jolla, CA, USA). The data are expressed as means ± SEM. Normality of data was evaluated with Shapiro–Wilk test. Student t-test or one-way ANOVA followed by Newman–Keuls post-hoc test was used for analysis. p values < 0.05 were considered statistically significant.

In order to assess the anti-inflammatory effect of FFAR agonists, we used RAW264.7 macrophages treated with LPS. We observed that FFAR1 agonist GW9508 (80 µM), FFAR3 agonist AR420626↗ (20 µM), and FFAR4 agonist GSK137647 (50 µM) significantly reduced the level of NO secreted to the culture medium without affecting cell viability (Figure 1A,B). The effect of FFAR agonists was more potent than the reference drug budesonide (BUD) at a concentration of 10 µM. Investigation into expression of FFAR1, FFAR2, FFAR3, and FFAR4 in macrophages revealed that treatment with LPS significantly downregulated FFAR2 and FFAR3 as compared to control (Figure 1C,D).

To further characterize the effect of FFAR agonists on the inflammatory response, we measured the secretion of IL-6 in cytokine (mixture of TNFα, IL-1β, IFNγ) and LPS-treated Caco-2 cells. First, we assessed the cytotoxicity of all compounds and selected concentrations that do not affect cell viability (Figure 2A). We found that only FFAR4 agonist GSK137647 (30 µM) significantly reduced IL-6 secretion in response to the mixture of cytokines. On the other hand, FFAR3 agonist AR420626↗ (50 µM) significantly enhanced secretion of IL-6 (Figure 2B). Investigation into expression of FFARs in Caco-2 cells revealed that treatment with cytokines downregulated FFAR1, FFAR2, and FFAR4 as compared to control (Figure 2C,D).

To evaluate the anti-inflammatory effect of FFAR agonists in the mouse GI tract, first, we used a well-established mouse model of GI inflammation induced by TNBS. The i.c. injection of TNBS resulted in reproducible colitis in mice, manifested by increased macroscopic colon-damage score and elevated MPO activity. We did not observe a significant decrease in macroscopic score, bowel thickness, and MPO activity after treatment with FFAR agonists. However, there was a clear, although not significant, anti-inflammatory effect of FFAR4 agonist GSK137647 (1 mg/kg bw, i.p., twice daily) (Figure 3A–C). Analysis of FFAR expression in the mouse colon revealed a significant increase of FFAR1 but not FFAR2–4 during inflammation (Figure 3D).

Additionally, we characterized the anti-inflammatory potential of FFAR agonists in the mouse model of UC induced by DSS. Animals treated with 4% DSS in drinking water developed severe colitis, characterized by increased macroscopic colon damage score, reduced colon weight, and elevated MPO activity (Figure 4).

The i.p. administration of FFAR4 agonist GSK137647 (1 mg/kg bw, twice daily) reversed colonic injury induced by DSS, as shown in Figure 4A–C.

Moreover, we observed that during DSS-induced colitis, expression of FFAR1 was significantly increased, while FFAR2 decreased in the colon (Figure 4D).

Analysis of FFAR expression in other parts of the GI tract revealed no changes in the tongue (Figure A1A), elevation of FFAR1 and decrease of FFAR3 and 4 in the esophagus (Figure A1B), increase of FFAR1, and decrease of FFAR3 in the stomach (Figure A1C), decrease of FFAR3 in the ileum (Figure A1D), and no alterations in the proximal colon (Figure A1E).

To further characterize the expression of FFAR in the healthy and inflamed colon, we used IHC staining. We observed that in healthy animals, FFAR1 and 2 are predominantly expressed in the colonic crypts, and this expression pattern was no longer visible after treatment with DSS, as the colonic architecture was highly disrupted (Figure 5A–D).

Moreover, in DSS-treated mice, FFAR1-positive cells were visible in both mucosal and submucosal layers of the colonic wall, whereas cells expressing FFAR2 tended to be closer to the luminal site of the colon (Figure 5B,D).

On the other hand, we observed higher number of FFAR3-positive cells in mucosal and submucosal layers of the inflamed colon as compared to the control, where FFAR3 was present in the colonic crypts (Figure 5E,F).

FFAR4 was detected in the colonic crypts of both control and DSS-treated animals, although the damage of mucosal architecture in the latter group caused entrapment of FFAR4-positive cells in the lumen of the disrupted crypts (Figure 5G,H).

Quantification of FFAR expression based on the IHC staining showed significantly decreased presence of FFAR1 and clear, although not significant, increase in the presence of FFAR3 in DSS-treated mice as compared to controls (Figure 5I).

To characterize the mechanism of anti-inflammatory action of FFAR4 agonist GSK137647 in the DSS model of colitis, we used a selective FFAR4 antagonist AH7614. As shown in Figure 6A, macroscopically, the action of FFAR4 agonist GSK137647 was reversed by the FFAR4 antagonist AH7614 (5 mg/kg bw, i.p.). Moreover, treatment with the antagonist blocked the beneficial effect of FFAR4 agonist GSK137647 on the relative colon weight (Figure 6B).

Administration of AH7614 caused increase in the MPO activity in the inflamed tissue compared with the control (Figure 6C).

First, we investigated the effect of FFAR agonists on cAMP-dependent ion transport in isolated mouse colon mounted in the Ussing chambers. Treatment with DSS caused reduction in the FSK-induced ΔIsc values as compared to controls (Figure 7).

FFAR1 agonist significantly decreased cAMP-dependent transport in control group but did not affect the inflamed mouse colon (Figure 7A).

Incubation with FFAR2 agonist significantly reduced ΔIsc in both healthy and DSS-treated mice (Figure 7B).

Adding FFAR3 agonist to the basolateral side of the colonic tissue only changed the ion flow in the samples from mice treated with DSS (Figure 7C).

FFAR4 agonist did not cause significant differences in either control or DSS groups (Figure 7D).

To examine ion flow dependence on voltage-dependent Na⁺ channel, we used VER. FFAR1 and FFAR2 agonists significantly lowered, while FFAR3 and FFAR4 ligands increased ion transport in control tissues challenged with VER (Figure 8). Incubation of inflamed tissue with FFAR1, FFAR2, and FFAR4 agonists significantly reduced the ion flow (Figure 8A,B,D), while FFAR3 increased this parameter (Figure 8C).

We used fluorescent probe (FITC-labeled dextran) to evaluate the paracellular permeability in mouse colon segments ex vivo. Permeability of colonic tissue to FITC-labeled dextran was increased by DSS at 60–120 min after the addition of the probe to the mucosal chamber. Treatment with FFAR1 agonist significantly weakened the effect of DSS on the permeability of the colonic tissue at 100 and 120 min after the addition of the FITC dextran (Figure 9A).

No differences caused by FFAR2 and FFAR3 agonists were observed (Figure 9B and Figure 2C).

FFAR4 had the strongest effect on the tissue, and the statistical differences with inflamed tissue were observed from 60 min (Figure 9D).

Next, we assessed TJs mRNA expression with qPCR using distal colon samples collected from control, DSS-treated, and DDS+FFAR4-treated mice. As shown in Figure 10, the expression of OCLN, CLDN1, CLDN2, CLDN7, and CLDN10 was changed in inflamed tissues in comparison to controls. FFAR4 restored the expression of OCLN, CLDN1, CLDN2, and CLDN10 (Figure 10A–C,G). Expression of CLDN7 was upregulated after treatment with FFAR4 agonist as compared to both control and DSS-treated animals (Figure 10F).

Data available, upon reasonable request, from the corresponding author.