Fig. 3

Iron supplementation differentially affectsmetabolic pathways and bacterial metagenomic contributionsin Apc^Min/+ mice receiving fecal microbiota transplantation from healthy controls (FMT-HC) and patients with colorectal cancer (FMT-CRC). A Heatmap of differential metabolic pathways significantly affected among FMT-HC and FMT-CRC mice fed iron sufficient (50 ppm) or iron excess (500 ppm) diets. Outliers, defined as 8 times the mean of the group, are indicated in black. Significantly affected pathways families exclusively in FMT-HC (P < 0.01) or FMT-CRC mice (P < 0.05) after FDR correction are framed in red. B, D Stacked bar charts showing the metagenomic contributors to the significantly affected metabolic pathways in B FMT-HC mice and D FMT-CRC mice fed the iron excess diet. C Proliferation of HT29 cells treated with culture broth (CTLR) or cell-free supernatant from B. pseudolongum cultures. n = 5 independent experiments. E Quantification of fecal butyrate, propionate, acetate, and isobutyrate in FMT-HC and FMT-CRC mice. Each symbol represents one mouse, FMT-HC (n = 24 (50 ppm); n = 26 (500 ppm)) and FMT-CRC (n = 25 (50 ppm); n = 27 (500 ppm)). P values were obtained using the generalized estimating equations (GEE) to correct for covariance structure of mice from a same donor (n = 3 mice/donor). F Assessment of HT29 cell proliferation in the presence of the SCFAs (butyrate, propionate and acetate, n = 5). P values were obtained using one-way ANOVA and post-hoc Dunnett test. Bars are means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, n.s.: non-significant