The Short Chain Fatty Acid Butyrate Imprints an Antimicrobial Program in Macrophages

Abstract

Host microbial cross-talk is essential to maintain intestinal homeostasis. However, maladaptation of this response through microbial dysbiosis or defective host defense toward invasive intestinal bacteria can result in chronic inflammation. We have shown that macrophages differentiated in the presence of the bacterial metabolite butyrate display enhanced antimicrobial activity. Butyrate-induced antimicrobial activity was associated with a shift in macrophage metabolism, a reduction in mTOR kinase activity, increased LC3-associated host defense and anti-microbial peptide production in the absence of an increased inflammatory cytokine response. Butyrate drove this monocyte to macrophage differentiation program through histone deacetylase 3 (HDAC3) inhibition. Administration of butyrate induced antimicrobial activity in intestinal macrophages in vivo and increased resistance to enteropathogens. Our data suggest that (1) increased intestinal butyrate might represent a strategy to bolster host defense without tissue damaging inflammation and (2) that pharmacological HDAC3 inhibition might drive selective macrophage functions toward antimicrobial host defense.

Graphical abstract

Figure 1

Increased Antimicrobial Activity by Macrophages Differentiated in Presence of Butyrate (A–H) Gentamicin protection assay on control macrophages and butyrate macrophages with a range of different bacteria. Macrophages were infected for 1 h with Salmonella enterica serovar Typhimurium (Salmonella) (A, E–H), adherent-invasive Escherichia coli (AIEC) (B), Staphylococcus aureus (S. aureus) (C) or Citrobacter rodentium (C. rodentium; D) followed by gentamicin treatment for 2 h before cell lysis. Values represent absolute CFU counts. Each dot is representative of one donor. Representative image of agar plate showing Salmonella CFU (A, right). (E) Gentamicin protection assay on macrophages treated with different SCFAs. (F) Kinetics of elimination of Salmonella by control macrophages and butyrate macrophages. (G) Short-term butyrate treatment: macrophages were treated for 3 h with butyrate prior to the gentamicin protection assay. (H) Butyrate macrophages were cultured in the absence of butyrate for 24 h prior to the gentamicin protection assay. Each dot represents one independent donor, experiments were repeated 3–8 times. Statistical significance was determined using Mann-Whitney U test ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001. Please also see Figure S1.

Figure 2

Increase of Antimicrobial LC3-Associated Immune Defense in Butyrate Macrophages (A) Extracellular acidification rate (ECAR) measured at steady state in control and butyrate macrophages. Data represent the mean of nine biological replicates from three independent experiments. (B–D) Quantification of glycolysis (B), glycolytic capacity (C) and glycolytic reserve (D). (E) Heatmap of metabolites that were significantly higher or lower in control and butyrate macrophages as detected by mass-spectrometry (left panel). Results are from five healthy donors. Right panel: Fold change of all significantly higher or lower metabolites in control macrophages (black closed circles) and butyrate macrophages (red closed circles). (F) AMPK phosphorylation (Thr172) measured by ELISA (n = 4 individual donors). (G) Percentage of pS6 (left) and representative blot of phosphorylation and quantification (right) of the ribosomal protein S6 and β-actin in control and butyrate macrophages at steady state. (H) Gentamicin protection assay performed on control macrophages, butyrate macrophages, and butyrate macrophages treated for 2 h with the mTOR activator MHY1485 (20 mM). (I) Representative immunoblot of the expression of LC3-II, P62, and β-actin at steady state or after 2 h infection with Salmonella. (J and K) Protein quantification performed by ImageJ of LC3-II (I) and P62 (J) compared to β-actin. (L) Degradation of GFP-Salmonella and LC3 induction was assessed by confocal microscopy. Representative images and quantification of GFP fluorescence and LC3 accumulation as outlined in Figure S7C. Data from six independent donors in two independent experiments. Scale bar 5 μm. Each dot represents one donor. Statistical significance was determined using Mann-Whitney U test ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001. Please also see Figures S2 and S3.

Figure 3

Butyrate Exposure Promotes an Antimicrobial Macrophage Phenotype (A) Single-cell RNA-seq analysis of control and butyrate treated macrophages (n = 5,836 cells) from two healthy donors. The t-SNE projection shows the five subpopulations of macrophages that were identified with a graph-based clustering algorithm (see Methods). (B) Quantitation of the proportions of control and butyrate macrophages ascribed to each of the sub-populations. Dashed lines indicate the sample pairs (shared donor identity). (C) Examples of the expression of marker genes identified (in both control and butyrate macrophages) for cells in each of the five clusters (Benjamini Hochberg [BH] adjusted p values < 0.05). (D) Changes in gene expression between control and butyrate macrophages. The genes shown in the heatmap are significantly differentially expressed (BH adjusted p value < 0.05, log₂ fold-change > 2) between control and butyrate macrophages in at least one of the four clusters of differentiated (non-cell-cycle associated) macrophages. (E) Selected examples of significantly enriched Gene Ontology (GO) categories among genes up (group i) or down (groups ii and iii) regulated by butyrate (BH adjusted p value < 0.05, Fishers exact test). Please also see Figure S4 and Tables S1–S3.

Figure 4

Upregulation of Calprotectin in Macrophages Differentiated in Presence of Butyrate (A) t-SNE plots of significantly (BH adjusted p value < 0.05) butyrate-induced antimicrobial genes (S100A8, S100A9, S100A12, LYZ, and FCN1) at the single-cell level. (B) Gene expression of S100A8, S100A9, S100A12 at steady state in control and butyrate macrophages. (C) Quantification of calprotectin in the supernatant by ELISA. (D) Intra-cellular expression of S100A8 and S100A9 protein at steady state by flow cytometry. (E) Representative immunoblot of calprotectin and β-actin at steady state from three donors. (F) Quantification of calprotectin expression at steady state. (G) Gentamicin protection assay against Salmonella performed on control and butyrate macrophages treated with scrambled siRNA or a mix of S100A8 and S100A9 siRNA. Each dot represents one donor. Statistical significance was determined using Mann-Whitney U test ^∗p < 0.05, ^∗∗p < 0.01 and ^∗∗∗p < 0.001. Please also see Figure S5.

Figure 5

Butyrate Promotes Antimicrobial Activity in Macrophages via HDAC Inhibitory Function (A–C) Gene expression of hcar2 (A), ffar2 (B), and ffar3 (C) on flow cytometry sorted CD19⁺ B, CD8⁺ T, naive CD4⁺ T, memory CD4⁺ T, CD56⁺ CD3⁻ NK cells, CD14⁺ monocytes, CD141⁺ DCs, and CD1c⁺ DCs from the blood of healthy donors. Each dot represents one donor. (D) Gentamicin protection assay on control and butyrate macrophages, or butyrate macrophages differentiated in the presence of butyrate with pertussis toxin (PT). (E) Protein expression of acetyled histone 3 (Ac-H3) and acetyled histone 4 (Ac-H4) in control and butyrate macrophages (data from 2 individual donors per condition). (F and G) Intra-cellular expression of tri-methylated lysine27 on histone 3 (3MeH3K27) (F) and acetylated lysine 27 on histone 3 (AcH3K27) (G) protein on control and butyrate macrophages by flow cytometry. (H, J, L, and N) Gentamicin protection assay on control macrophages, butyrate macrophages, and macrophages differentiated in the presence of valproate, phenylbutyrate (H), SAHA (J), TMP195, and tubacin (L), SBHA, 1-naphthohydroxamic acid (NA), or RGFP966 (N). (I, K, M, and O) Gene expression of S100A8 in control macrophages, butyrate macrophages, and macrophages differentiated in the presence of valproate, phenylbutyrate (I), SAHA (K), TMP195, and tubacin (M), SBHA, 1-naphthohydroxamic acid (NA), or RGFP966 (O). (P) Gentamicin protection assay in control and butyrate macrophages with and without HDAC3 siRNA-mediated gene silencing. (Q) Extracellular acidification rate in control and butyrate macrophages, as well as in RGFP966 and TMP195 treated macrophages. Data represent the mean of three biological replicates. For the gentamicin protection assay, the percentage of CFU (of control) was calculated from mean value for control group. Each dot represents one independent donor. For pairwise comparison Mann-Whitney U test was performed and for multiple group comparisons a one-way ANOVA (Kruskal-Wallis test) was performed. ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001. Please also see Figure S5E.

Figure 6

Induction of Antimicrobial Activity by Butyrate in Macrophages In Vivo (A) WT mice received sodium butyrate in drinking water (150 mM final concentration) or PBS control for 7 days. At day 7, colonic segments were digested, macrophages were isolated by flow cytometry sorting and a gentamicin protection assay was performed. Each dot represents macrophages pooled from ten mice. Four independent experiments are shown. (B) Mouse bone marrow progenitor cells from mice gavaged with sodium butyrate or with PBS were differentiated into macrophages in the presence of M-CSF or with M-CSF with butyrate as a positive control. A gentamycin assay was performed at day 7 of differentiation. (C and D) Mice received butyrate or PBS 5 days prior to oral infection with Salmonella typhymurium def aroA (1 × 10⁹ bacteria/mouse). 2 days post-infection bacterial dissemination was assessed in MLN, spleen, liver (C), and caecum (D). Each dot represents a mouse. (E) Colitis score of control and butyrate-treated mice either uninfected or infected with Salmonella. (F) Representative H&E stained colon sections from control and butyrate-treated mice either uninfected or infected with Salmonella (original magnification 100x). (G and H) Mice were treated with 150 mM sodium butyrate or with PBS 3 days prior and every other day after oral infection with Citrobacter rodentium (1 × 10⁹ bacteria/mouse). Mice were weighed daily. Lines represents mean of 3 mice (G). At day 7 post infection bacterial dissemination was assessed in the spleen and in the liver (H). Each dot represents a mouse. Statistical significance was determined using Mann-Whitney U test ^∗p < 0.05, ^∗∗p < 0.01, and ^∗∗∗p < 0.001. Please also see Figure S6.