SCFAs induce mouse neutrophil chemotaxis through the GPR43 receptor

Abstract

Short chain fatty acids (SCFAs) have recently attracted attention as potential mediators of the effects of gut microbiota on intestinal inflammation. Some of these effects have been suggested to occur through the direct actions of SCFAs on the GPR43 receptor in neutrophils, though the precise role of this receptor in neutrophil activation is still unclear. We show that mouse bone marrow derived neutrophils (BMNs) can chemotax effectively through polycarbonate filters towards a source of acetate, propionate or butyrate. Moreover, we show that BMNs move with good speed and directionality towards a source of propionate in an EZ-Taxiscan chamber coated with fibrinogen. These effects of SCFAs were mimicked by low concentrations of the synthetic GPR43 agonist phenylacetamide-1 and were abolished in GPR43(-/-) BMNs. SCFAs and phenylacetamide-1 also elicited GPR43-dependent activation of PKB, p38 and ERK and these responses were sensitive to pertussis toxin, indicating a role for Gi proteins. Phenylacetamide-1 also elicited rapid and transient activation of Rac1/2 GTPases and phosphorylation of ribosomal protein S6. Genetic and pharmacological intervention identified important roles for PI3Kγ, Rac2, p38 and ERK, but not mTOR, in GPR43-dependent chemotaxis. These results identify GPR43 as a bona fide chemotactic receptor for neutrophils in vitro and start to define important elements in its signal transduction pathways.

Conflict of interest statement

Competing Interests: Mohammad Bohlooly-Y is from AstraZeneca Transgenic and Comparative Genomic R&D Mölndal. This does not alter the authors' adherence to all the PLoS ONE policies on sharing data and materials.

Figures

Figure 1. SCFAs and phenylacetamide induce neutrophil migration in a transwell chamber.

Chemotaxis of BMN to the indicated concentration of propionate (A), butyrate (B), acetate (C), phenylacetamide (p-acetamide) (D) and serine (E) was evaluated in a transwell migration assay. The level of migration to 1 µM fMLP is shown (filled square in A, note scale). Data show the mean ± S.E.M. (n = 3) number of migrating neutrophils normalized to the control condition (non-treated). *p<0.05 compared with control condition (one-way ANOVA with Dunnet's post test).

Figure 2. SCFAs induce neutrophil chemotaxis in the EZ-Taxiscan Chamber.

Centre-zeroed tracks of purified BMN in an EZ-Taxiscan Chamber migrating towards the indicated concentrations of propionate, phenylacetamide (p-acetamide) or fMLP (the source is at the bottom of the diagram, scale is in µm) (A). The vector of each cell was measured as it crossed a horizon of 50 µm and these vectors are displayed as a horizon plot. Each 18° segment of the horizon plot indicates the proportion of cells that were moving within each vector. When the population was significantly moving in one direction (Raleigh test for unimodal clustering, p<0.05) the mean vector (arrow) and 95% confidence limit (shaded area) is shown. The tracks presented in A were analysed to show the percentage of motile cells (cells crossing a 50 µm horizon) (B), migratory index (of the motile cells) (C) and speed (of the motile cells) (D) (mean ± S.E.M.). *p<0.05 compared with control condition (one-way ANOVA with Dunnet's post test). The data in A–D are derived from 8–10 movies.

Figure 3. SCFAs increase neutrophil spreading on a fibrinogen matrix.

Purified BMN were mixed with PBS or the indicated concentrations of agonists and immediately applied to fibrinogen coated plates for 15 or 30 min. Adherent cells were visualized by phase contrast microscopy (32×, Zeiss Axiovert) and the number of adherent (A) or spread (B) neutrophils was then quantitated (7 random fields per replicate). Data are mean ± S.E.M. (n = 3). *p<0.05 compared with control condition (one-way ANOVA with Dunnet's post test).

Figure 4. GPR43−/− neutrophils do not migrate in response to propionate or phenylacetamide, but display a normal response to fMLP.

Chemotaxis of wild type (WT) and GPR43−/− neutrophils was evaluated in a transwell migration assay (A) and EZ-Taxiscan Chamber (B–C). Chemotaxis of wildtype or GRR43−/− neutrophils in a transwell migration assay was measured to propionate (1 mM), phenylacetamide (p-acetamide, 10 µM) or fMLP (1 µM). Data show the mean ± S.E.M. (n = 3) number of migrating neutrophils normalized to the control condition (non-treated). *p<0.05 compared to wildtype cells (paired t-test) (A). Chemotaxis of wildtype or GRR43−/− neutrophils in an EZ-Taxiscan chamber was measured to propionate (25 mM), phenylacetamide (p-acetamide, 50 µM) or fMLP (3 µM). The mean ± S.E.M. (n = 3) % of neutrophils migrating (past a 50 µm horizon) *p<0.05 compared with control condition (paired t-test) (B) and centre-zeroed tracks and horizon plots, as described in figure 1, (C) are shown. The data in B–C are derived from 6–10 movies.

Figure 5. Propionate and phenylacetamide activate PKB, ERK1/2, p38 and Rac1/2 through GPR43 and Gi.

Purified BMN were incubated with PBS, propionate (10 mM), phenylacetamide (25 µM) or fMLP (10 µM) for the indicated times. Whole-cell lysates were prepared and analyzed by immunoblotting using specific Abs to phosphorylated forms of PKB, ERK1/2, p38 (Ai) or the S6 ribosomal protein (Aii). The cytosolic proteins β-cop or β-actin served as loading controls. Activation of PKB, ERK1/2 and p38 by phenylacetamide (25 µM for 60 seconds; PATM) was also quantified by densitometry in neutrophils isolated from GPR43−/− mice (Bi) or neutrophils pre-treated with pertussis toxin (PTX) (2 µg/mL) (Bii) and are displayed as mean ± S.E.M. (n = 4–5) relative to the wildtype, fMLP-treated, sample. The dashed line indicates the basal level of phosphorylation in wildtype cells (Bi) or untreated (Bii) cells. In order to assess Rac1 and Rac2 activation, purified BMN were incubated with phenylacetamide (25 µM) or PBS for the indicated time and lysates subjected to GST-PAK-CRIB pull down. A representative blot is shown (Ci) along with quantified data from 4 independent experiments (Cii) (mean ± S.E.M). *p<0.005 compared to PBS control [paired t-test]) for Rac1 and for Rac2.

Figure 6. Neutrophil chemotaxis in response to propionate, phenylacetamide or fMLP requires PI3Kγ, MAPK, Rac2 but not mTor.

Chemotaxis of wildtype BMN pre-incubated with the indicated inhibitor (200 nM wortmaninn, 1 µM PD184352 or 10 µM SB203580) (Ai), or derived from wildtype (WT) or p110γ−/− mice (B) was assessed in response to 25 mM propionate, 50 µM phenylacetamide (P-acetamide) or 3 µM fMLP in an Ez-Taxiscan chamber. Data are expressed as the mean ± S.E.M. (n = 3) % of neutrophils migrating. *p<0.05 compared with control condition (cells pretreated with DMSO) (one-way ANOVA with Dunnet's post test) (Ai), *p<0.05 compared to the wildtype (paired t-test) (B). In addition, chemotaxis of wildtype BMN pre-incubated with the indicated inhibitor (100 nM rapamycin or 100 nM torin) (Aii), or derived from wildtype (WT) or Rac2−/− neutrophils (C) towards 1 mM propionate, 10 µM phenylacetamide or 1 µM fMLP was assessed in a transwell chemotaxis assay and is expressed as the mean ± S.E.M. of migrating neutrophils normalized to the control condition (non-treated). The data show a representative result of 2 experiments each performed in triplicate. *p<0.05 compared to the wildtype (paired t-test).