Timing of host food intake and microbial activity are linked.

(A) Cell counts per gram feces in the EAM mouse (n=8) at the end of the light cycle (zt=12h), measured by qPCR. (B) Food intake and hydrogen production of the EAM mouse over time. Two days of measurement are shown, measurements are taken every 24min (n=16). Curves were calculated using the geom_smooth function of the package ggplot2 in R with standard settings. SE is illustrated with grey shading surrounding the trend line. (C) Food intake and hydrogen production of the germ-free mouse over time (n=8).

Lactulose treatment disturbs the diurnal rhythm of microbial metabolism.

(A) Hydrogen production increases after lactulose treatment (purple, n=8), but not after treatment with PBS (orange, n=8). Arrows at zt=3h indicate treatment. An unpaired t-test was performed on final hydrogen measurements (p>0.0001). (B) Hydrogen production after lactulose treatment (purple, n=6) is temporary and reverses back to normal levels after treatment, and is not observed in PBS treatment (orange, n=5). (C) Bacterial counts in cecum 5h after lactulose (purple, n=12) and PBS treatment (orange, n=10). E. rectale population increases in lactulose treated mice compared to PBS treated mice (p<0.05), B. theta and E. coli counts do not change significantly. (D) Cecum pH of 3MM (n=10 for lactulose, n=7 for PBS treatment) and germ-free (n=8 in each group) mice 5h after treatment. The pH decreases after lactulose treatment in 3MM (p<0.0001, unpaired t-test) and germ-free (p<0.05) mice. (E) Concentrations of microbial metabolites in cecum content of 3MM mice 5h after treatment with lactulose (purple, n=10) or PBS (orange, n=7). Acetate, butyrate, and succinate concentrations increase in lactulose treated as compared to PBS treated mice (**, ***, **** indicate significance levels smaller than 0.01, 0.001, and 0.0001, respectively); unpaired t-test, Benjamini-Hochberg corrected for multiple testing. (F) Concentrations of microbial metabolites in cecum content of germ-free mice 5h after treatment with lactulose (purple, n=8) or PBS (orange, n=8). There is no significant difference between groups (unpaired t-test, Benjamini-Hochberg corrected for multiple testing).

Lactulose feeding acutely alters clock gene expression in small intestinal tissue.

(A) Schematic of the clock gene network in mice. (B) Clock gene expression in ileum tissue in 3MM (N=10 for lactulose treatment, n=7 for PBS treatment) and germ-free (n=8 in each group) mice 5h after treatment, measured via qPCR. Fold change values are relative to the average of the respective PBS control groups. Statistical significance was tested using upaired t-tests and Benjamini-Hochberg corrected for multiple testing (* and ** indicate p<0.05 and p<0.01, respectively). (C) Clock gene expression in ileum tissue of 3MM mice during 24h after treatment (treatment starts at time 0h). All gene expression data was normalized to samples collected from untreated mice at zt3 (control). Expression differed significantly between treatment and control groups for cry1, cry2, and per2 (two-way ANOVA (time, treatment), p<0.05).

Lactulose treatment decreases host food intake in the following active cycle.

(A) Food intake rate in 3MM mice in the dark phase prior to treatment (control) and following treatment (treatment), after lactulose (n=8) and PBS (n=7) treatment. Food intake is significantly reduced after lactulose treatment (paired t-test, p<0.01). (B) Food intake rate in germ-free mice after lactulose (n=5) and PBS (n=4) treatment. Food intake is not significantly different between groups (paired t-test). (C) Food intake rate of 3MM mice during the dark phase prior to treatment with fermentation products (control) and the dark phase after treatment (treatment). Food intake is significantly reduced after treatment (paired t-test, p<0.01). Treatment was done by oral gavage of 66mg of fermentation product in 100µL of PBS, with a 4:3:3 ratio of sodium succinate, sodium acetate, and sodium butyrate, respectively. (D) 3MM mouse total calorie intake after accounting for the calories directly provided by fermentation products in treatment. Calory intake remains significantly reduced after treatment when calories provided by the treatment are considered (paired t-test, p<0.05). (E) 3MM mouse total calorie intake after accounting for calories provided by the microbiota after lactulose metabolism. Calory intake remains significantly reduced after lactulose treatment when calories provided by microbial metabolism are taken into account (paired t-test, p<0.05).

E. rectale is the principal hydrogen producer in the 3MM mirobiota.

(A) Microbial cell counts based on qPCR in 3MM mouse feces. Fecal pellets were sampled over 2 days in 8 3MM mice. (B) Food intake and hydrogen production in 3MM mice (n=16) over two days, with measurements taken every 24min. Trend lines were plotted using the geom_smooth function in the ggplot2 package in R with standard settings, shaded areas around the trend lines represent standard error of the mean. (C) Food intake and hydrogen production of gnotobiotc mice colonized with B. theta and E. coli only (n=4). In all panels, shaded background indicates dark periods in the animal facility, from 6pm (zt=12) to 6am (zt=24).

Lactulose changes the diurnal pattern of microbial metabolism.

(A) Hydrogen production temporarily increases after treatment with lactulose (n=5), but not after treatment with PBS (n=6), in SPF mice. Treatment is indicated by a black arrow. (B) Production of fermentation products by the bacterial strains constituting the 3MM microbiota grown in axenic culture in media with lactulose as carbohydrate source. Measurements were taken in stationary phase (*, **, ***, ****, indicate p<0.05, 0.01, 0.001, and 0.0001, respectively; ANOVA with Tukey’s post-hoc test). (C) Fecal water content in 3MM mice treated with lactulose (n=8) and PBS (n=7), and germ-free mice treated with lactulose (n=4) and PBS (n=4). The water content decreases significantly upon lactulose treatment in 3MM mice (p<0.01, unpaired t-test). (D) Estimated production of total fecal wet weight during 5h in 3MM mice after treatment with lactulose (n=4) and PBS (n=4), and in germ-free mice treated with lactulose (n=4) and PBS (n=4). This data was calculated by experimentally measuring total fecal dry weight and correcting it with the fecal water content determined for the respective microbiota and treatment in (C). Output of fecal wet weight increases significantly in 3MM mice after lactulose treatment (unpaired t-test, p<0.05). (E) Water content of cecum content in 3MM mice treated with lactulose (n=10) and PBS (n=7). (F) Cecum weight of 3MM and germ-free mice treated with PBS or lactulose. Germ-free mice have significantly higher cecum weights than 3MM mice (p<0.01, unpaired t-test), which already have higher cecum weights than conventionally colonized mice (around 0.5g)(35).

Even higher dose lactulose treatment does not lead to an osmotic effect in 3MM mice, but does so in SPF mice.

(A) Water content in fresh fecal pellets collected in regular intervals after PBS or lactulose treatment, in 3MM and SPF mice. Lines indicate linear regression on the data. Fecal water content does not increase after PBS treatment in 3MM or SPF mice. Lactulose treatment does also not lead to an increase in water content in 3MM mice, but did lead to an increase in SPF mice, indicating that the osmotic effect of lactulose depends on the colonization state of the animals. (B) Cecum water content 5h after treatment. Only lactulose treated SPF mice showed a significant increase in water content in cecum content 5h after treatment (p<0.01, ANOVA with Tukey’s post-hoc test). (C) Total fecal output during 5h after treatment was not significantly different between groups (ANOVA). This data was collected the same way as in Figure S2D, by collecting all fecal pellets from cage bedding, measuring their dry weight, and correcting it by the mean water content of the fresh feces in the respective group.

Targeted induction of microbial metabolism by lactulose treatment acutely alters clock gene expression in small intestinal tissue.

(A) Clock gene expression measured by qPCR in tissue samples of ileum, cecum, and colon of 3MM mice after treatment with lactulose (n=10) and PBS (n=7). Significant differences in gene expression between groups were detected in samples from the ileum (unpaired t-test, multiple testing correction). (B) Clock gene expression in tissue samples of ileum, cecum, and colon of germ-free mice after lactulose (n=8) and PBS (n=8) treatment. No significant differences were found in gene expression levels between treatment groups (unpaired t-test, multiple testing correction). (C) Clock gene expression in ileum tissue of SPF mice after treatment with lactulose (n=5) or PBS (n=5). Cry1 expression was significantly different between treatment groups (unpaired t-test, Benjamini-Hochberg correction). In all panels, * and ** indicate p<0.05 and p<0.01, respectively.

Induction of microbial metabolism transiently decreases host food intake in the following active cycle.

(A) Total food intake in 3MM mice before treatment (control) and after lactulose (n=8) and PBS (n=7) treatments. Food intake was significantly lower after lactulose treatment (paired t-test, p<0.01). (B) Food intake rate in 3MM mice in the dark phase before treatment (control) and the two dark phases following treatment (treatment, treatment+1). The food intake rate is not significantly different between the days (ANOVA). (C) Total food intake in 3MM mice before treatment (control) and in the two dark cycles after treatment with lactulose (n=3) and PBS (n=4), when treated at 8:30pm (zt=14.5, 2.5h into the dark cycle). No significant differences were found between groups (ANOVA). (D) Food intake rate for the experiment shown in in (C). (E) Total food intake in germ-free mice in the dark cycle before treatment (control) and the dark cycle after treatment with lactulose (n=7) and PBS (n=7). No significant differences were found between the days (paired t-test). (F) Total food intake of SPF mice before treatment (control) and in the two dark cycles after treatment with lactulose (n=4) and PBS (n=4). No significant differences were found between the days (ANOVA).

Clock gene expressing in perpheral tissues and food intake rates in SPF mice after PBS or lactulose treatment.

(A) Clock gene expression in hypothalamus and liver of EAM mice after lactulose (n=5) and PBS (n=4) treatment, measured via qPCR. Gene expression levels are not significantly different between groups (unpaired t-test). (B) Food intake rates of SPF mice in the dark phase before treatment (control) and the dark phase following treatment with fermentation products. Food intake rates did not differ significantly between the days (paired t-test).

Concentration measurements of metabolic hormones after treatment with PBS or lactulose.

(A) Total GLP1 concentration in portal vein plasma 5h after lactulose or PBS treatment in 3MM, germ-free, and SPF mice. We found no significant differences between groups (unpaired t-test). (B) PYY concentration in portal vein plasma after lactulose or PBS treatment in 3MM, germ-free, and SPF mice. Groups were not significantly different (unpaired t-test). (C) Total Ghrelin concentration in serum from cardiac puncture after lactulose or PBS treatment in 3MM and germ-free mice. No significant differences were found between groups in 3MM mice. Ghrelin was significantly increased (p<0.05) in germ-free mice after lactulose treatment (unpaired t-test). (D) Total Leptin concentration in serum from cardiac puncture after lactulose or PBS treatment in 3MM and germ-free mice. Concentrations were not significantly different between groups (unpaired t-test).