Volume 138, Issue 5, Pages 1772-1782.e4 (May 2010) Short-Chain Fatty Acids Regulate the Enteric Neurons and Control Gastrointestinal Motility in Rats  Rodolphe Soret, Julien Chevalier, Pierre De Coppet, Guillaume Poupeau, Pascal Derkinderen, Jean Pierre Segain, Michel Neunlist  Gastroenterology  Volume 138, Issue 5, Pages 1772-1782.e4 (May 2010) DOI: 10.1053/j.gastro.2010.01.053 Copyright © 2010 AGA Institute Terms and Conditions

Figure 1 Resistant starch diet (RSD) increased the proportion of ChAT-immunoreactive (IR) myenteric neurons in the rat. (A) Triple immunohistochemical labeling of myenteric neurons of proximal colon from rats fed a standard diet (SD; upper row) or a RSD (lower row) for 14 days. Myenteric neurons were identified with anti-Hu antibody (left column) and stained with anti-ChAT (middle column, arrowhead) and anti-nNOS (right column, arrow) antibodies. (B) Quantitative analysis revealed that the proportion of ChAT-IR neurons (normalized to Hu-IR neurons) was significantly increased with RSD as compared to SD (n = 6; *P < .028, t test). (C) The proportion of nNOS-IR was similar in both diets. Scale bar, 50 μm. Data are presented as the mean ± SEM. Gastroenterology 2010 138, 1772-1782.e4DOI: (10.1053/j.gastro.2010.01.053) Copyright © 2010 AGA Institute Terms and Conditions

Figure 2 Butyrate increased the proportion of ChAT-immunoreactive (IR) myenteric neurons in vivo. (A) Triple immunohistochemical labeling of myenteric neurons in proximal colon in rat following 48 hours of intracaecal perfusion of saline solution (left column) or 10 mmol/L butyrate (right column) solutions. Myenteric neurons were identified with anti-Hu antibody (upper row) and stained with anti-ChAT (middle row, arrowhead) and anti-nNOS (lower row, arrow) antibodies. (B) Quantitative analysis revealed that butyrate treatment dose-dependently increased the proportion of ChAT-IR neurons (n = 6; *P < .05, one-way ANOVA). (C) The proportion of nNOS-IR was not modified by butyrate. (D) Intracaecal perfusion of butyrate (5 mmol/L) specifically increased the proportion of ChAT-IR neurons among total Hu-IR neurons as compared to acetate (5 mmol/L) or propionate (5 mmol/L) and to control (n = 6; *P < .05, one-way ANOVA). (E) The proportion of nNOS-IR neurons was not modified by any short chain fatty acid. Scale bar, 25 μm. Data are presented as the mean ± SEM. Gastroenterology 2010 138, 1772-1782.e4DOI: (10.1053/j.gastro.2010.01.053) Copyright © 2010 AGA Institute Terms and Conditions

Figure 3 Butyrate directly increased the proportion of ChAT-immunoreactive (IR) neurons in primary culture. (A) Triple immunohistochemical labeling of primary culture of ENS untreated (left column) or treated with butyrate (0.5 mmol/L) (right column). Neurons were identified with anti-Hu antibody (upper row) and stained with anti-ChAT (middle row, arrowhead) and anti-nNOS (lower row, arrow) antibodies. (B) Quantitative analysis revealed that the proportion of ChAT-IR neurons (normalized to the number of Hu-IR neurons) was significantly increased following butyrate treatment at 0.5 mmol/L and 1 mmol/L, as compared to control (n = 6; *P < .01, one-way ANOVA). (C) In contrast, the proportion of nNOS-IR neurones was not modified by butyrate except at 1 mmol/L butyrate concentration which reduced the proportion of nNOS-IR neurons. (D) Butyrate treatment (0.5 mmol/L) time-dependently increased the proportion of ChAT-IR neurons (n = 6; *P < .001, one-way ANOVA). (E) Treatment with 0.5 mM butyrate significantly increased ChAT mRNA expression as early as 6 hours after addition of butyrate (n = 6; *P < .05 as compared t = 0, one-way ANOVA). Scale bar, 25 μm. Data are presented as the mean ± SEM. Gastroenterology 2010 138, 1772-1782.e4DOI: (10.1053/j.gastro.2010.01.053) Copyright © 2010 AGA Institute Terms and Conditions

Figure 4 Characterization of monocarboxylate transporters expression in primary culture of enteric nervous system (ENS). (A) Quantitative PCR analysis of MCT1, 2, 3, 4 mRNA expression in primary culture of ENS showed a significantly larger expression of MCT2 than of other transporters (n = 11; *P < .01, compared to MCT1, 3, 4; #P < .01, compared to MCT1, one-way ANOVA). (B) Immunostaining of primary culture of ENS showed that Hu-immunoreactive (IR) neurons (upper panel) expressed MCT2 (lower panel, arrowhead). (C) Following, transfection of primary culture of ENS with mock siRNA a large number of Hu-IR neurons (upper left panel) expressed MCT2 (lower left panel, arrowhead). In contrast, after transfection with MCT2 siRNA, a few number of Hu-IR neurons (upper right panel) were MCT2 (lower right panel, arrowhead). (D) Quantitative analysis showed that butyrate significantly increased the proportion of ChAT-IR neurons in cells treated with mock siRNA. This proportion was significantly reduced in cells treated with MCT2 siRNA (n = 6; *P < .01 compared to control without butyrate, #P < .01 compared to control with butyrate one-way ANOVA). Scale bar, 25 μm. Data are presented as the mean ± SEM. Gastroenterology 2010 138, 1772-1782.e4DOI: (10.1053/j.gastro.2010.01.053) Copyright © 2010 AGA Institute Terms and Conditions

Figure 5 Butyrate induced increased in the proportion of ChAT-IR neurons involves Src-kinase signalling pathway. (A) Immunostaining of primary culture of ENS treated for 24 hours with 0.5 mmol/L butyrate in the absence or presence of the Src kinase inhibitor (PP2 or SU6656). Neurons were identified with anti-Hu antibody (upper row) and stained with anti-ChAT antibody (lower row). In presence of butyrate a large number of Hu-IR neurons (upper left panel) were ChAT-IR (lower left panel). However in presence of butyrate and PP2, a few number of Hu-IR neurons (upper middle panel) were ChAT-IR (lower middle panel). Similarly, in presence of another Src kinase inhibitor (SU6656) and butyrate, a few number of Hu-IR neurons (upper right panel) were also ChAT-IR (lower right panel). (B) Quantitative analysis revealed that the effects of butyrate upon the proportion of ChAT-IR neurons (normalized to the total neuronal population identified with Hu) was significantly reduced in presence of PP2 and SU6656 (n = 6; *P < .01, one-way ANOVA). (C) Quantitative analysis revealed that the effects of butyrate upon nNOS-IR (normalized to the total neuronal population identified with Hu) was not reduced in presence of PP2 and SU6656 (n = 6; *P < .01, one-way ANOVA). Scale bar, 25 μm. Data are presented as the mean ± SEM. Gastroenterology 2010 138, 1772-1782.e4DOI: (10.1053/j.gastro.2010.01.053) Copyright © 2010 AGA Institute Terms and Conditions

Figure 6 Butyrate increased in vivo colonic transit and ex vivo neurally mediated contractile response. Rats fed with RSD-supplemented diet showed an increased colonic transit as assessed by measurement of glass bead expulsion time in comparison to SD (A; n = 6; *P < .05, t test). Ex vivo measurements of contractile activity were performed in organ chamber. Quantitative analysis of the AUC (measured during the EFS) showed a larger AUC in butyrate-treated tissue compared with the control (B; n = 5; #P < .05 compared with control, t test). Addition of L-NAME (5·10−5 mol/L) did not significantly modify the AUC in both conditions. In contrast, in the presence of atropine (10−6 mol/L) the AUC was significantly decreased in both conditions (B; n = 5; *P < .05 compared with L-NAME, t test). The atropine-sensitive AUC (ΔAUC; difference between AUC in the presence of L-NAME and atropine and AUC in the presence of L-NAME) was significantly larger in the presence of butyrate compared with control (C; n = 5; P < .05, paired t test). Data are presented as the mean ± SEM. Gastroenterology 2010 138, 1772-1782.e4DOI: (10.1053/j.gastro.2010.01.053) Copyright © 2010 AGA Institute Terms and Conditions

Supplementary Figure 1 Specificity of the ChAT antibody. Immunohistochemical staining was performed in primary culture of ENS. (A) The primary ENS culture was stained with the goat anti-ChAT antibody alone. Under these conditions, no labeling of ChAT-IR neurons was observed. (B) To show the specificity of ChAT antibody, the antibody was first preincubated with the ChAT antigen (3 hours; 1:100; Millipore). Next, the primary ENS culture was incubated with the antibody-antigen solution for 3 hours and then stained with the secondary antibody (anti-goat cyanine 3 [Cy3]). Under these conditions, no ChAT-IR neuron was stained. (C) The primary ENS culture was stained sequentially with the goat-anti ChAT antibody and the secondary anti-goat Cy3 antibody. We then observed a strong ChAT-IR staining in enteric neurons (C). Scale bar, 25 μm. Gastroenterology 2010 138, 1772-1782.e4DOI: (10.1053/j.gastro.2010.01.053) Copyright © 2010 AGA Institute Terms and Conditions

Supplementary Figure 2 Butyrate induced acetylation of histone H3 in primary culture of ENS. A few neurons of primary culture ENS identified with Hu (A) were IR for histone H3 lysine 9 (H3K9) (C). After 24 hours of treatment with butyrate (500 μmol/L), a large number of Hu-IR neurons (B) were also H3K9-IR (D). After 24 hours of incubation with TSA (0.25 μmol/L), an inhibitor of HDAC, a large number of Hu-IR neurons (E) were also H3K9-IR (F). TSA significantly increased the proportion of ChAT-IR neurons (G) (normalized to the total neuronal population identified with Hu; n = 6; *P < .001, t test). In the absence of butyrate, few neurons identified with Hu (H) were also H3K9-IR (I). After 24 hours of treatment with butyrate (0.5 mmol/L), a large number of Hu-IR neurons (J) were also H3K9-IR (K). In the presence of Src kinase inhibitor (PP2) and butyrate, a few Hu-IR neurons (L) were also H3K9-IR (M). Similarly, in the presence of another Src kinase inhibitor (SU6656) and butyrate, a few Hu-IR neurons (N) were also H3K9-IR (O) (n = 6; *P < .001, t test). Scale bar, 50 μm (A–F) and 100 μm (H–O). Data are presented as the mean ± SEM. Gastroenterology 2010 138, 1772-1782.e4DOI: (10.1053/j.gastro.2010.01.053) Copyright © 2010 AGA Institute Terms and Conditions