Steamed broccoli sprouts alleviate DSS-induced inflammation and retain gut microbial biogeography in mice

ABSTRACT Inflammatory bowel diseases (IBDs) are devastating conditions of the gastrointestinal tract with limited treatments, and dietary intervention may be effective and affordable for managing symptoms. Glucosinolate compounds are highly concentrated in broccoli sprouts, especially glucoraphanin (GLR), and can be metabolized by certain mammalian gut bacteria into anti-inflammatory isothiocyanates, such as sulforaphane. Gut microbiota exhibit biogeographic patterns, but it is unknown if colitis alters these or whether the location of glucoraphanin-metabolizing bacteria affects anti-inflammatory benefits. We fed specific pathogen-free C57BL/6 mice either a control diet or a 10% steamed broccoli sprout diet and gave a three-cycle regimen of 2.5% dextran sodium sulfate (DSS) in drinking water over a 34-day experiment to simulate chronic, relapsing ulcerative colitis (UC). We monitored body weight, fecal characteristics, lipocalin, serum cytokines, and bacterial communities from the luminal- and mucosal-associated populations in the jejunum, cecum, and colon. Mice fed the broccoli sprout diet with DSS treatment performed better than mice fed the control diet with DSS, and had significantly more weight gain, lower Disease Activity Index scores, lower plasma lipocalin and proinflammatory cytokines, and higher bacterial richness in all gut locations. Bacterial communities were assorted by gut location but were more homogenous across locations in the control diet + DSS mice. Importantly, our results showed that broccoli sprout feeding abrogated the effects of DSS on gut microbiota, as bacterial richness and biogeography were similar between mice receiving broccoli sprouts with and without DSS. Collectively, these results support the protective effect of steamed broccoli sprouts against dysbiosis and colitis induced by DSS. IMPORTANCE Evaluating bacterial communities across different locations in the gut provides a greater insight than fecal samples alone and provides an additional metric by which to evaluate beneficial host-microbe interactions. Here, we show that 10% steamed broccoli sprouts in the diet protects mice from the negative effects of dextran sodium sulfate-induced colitis, that colitis erases biogeographic patterns of bacterial communities in the gut, and that the cecum is not likely to be a significant contributor to colonic bacteria of interest in the DSS mouse model of ulcerative colitis. Mice fed the broccoli sprout diet during colitis performed better than mice fed the control diet while receiving DSS. The identification of accessible dietary components and concentrations that help maintain and correct the gut microbiome may provide universal and equitable approaches to IBD prevention and recovery, and broccoli sprouts represent a promising strategy.

Important features (SVs) were identified through permutational random forest analysis, and only the features important to this group (>50 reads) are listed out of 188 significant (p < 0.05) features across all treatments.Model accuracy was 98%.Bacterial sequence variants (SV) are identified as the lowest level of taxonomic identity possible, with "NA" indicating which could not be identified to species, and the number indicating which specific SV it was.Four treatment groups were used in a 34-day chronic relapsing model of colitis: control diet, control diet with DSS added to drinking water, control diet adjusted with 10% by weight steamed broccoli sprouts, and 10% broccoli sprout diet with DSS added to drinking water (Figure 1).Bacterial communities were sampled from several locations in the gastrointestinal tract at the end of the study.

Figure S3. Abundance of bacterial sequence variants identified as differential for the Control+DSS group.
Important features (SVs) were identified through permutational random forest analysis, and only the features important to this group (>50 reads) are listed out of 188 significant (p < 0.05) features across all treatments.Model accuracy was 98%.Bacterial sequence variants (SV) are identified as the lowest level of taxonomic identity possible, with "NA" indicating which could not be identified to species, and the number indicating which specific SV it was.Four treatment groups were used in a 34-day chronic relapsing model of colitis: control diet, control diet with DSS added to drinking water, control diet adjusted with 10% by weight steamed broccoli sprouts, and 10% broccoli sprout diet with DSS added to drinking water (Figure 1).Bacterial communities were sampled from several locations in the gastrointestinal tract at the end of the study.

Figure S4
. Abundance of bacterial SVs identified as differential for the 10% steamed Broccoli Sprouts group.Important features (SVs) were identified through permutational random forest analysis, and only the features important to this group (>50 reads) are listed out of 188 significant (p < 0.05) features across all treatments.Model accuracy was 98%.Bacterial sequence variants (SV) are identified as the lowest level of taxonomic identity possible, with "NA" indicating which could not be identified to species, and the number indicating which specific SV it was.Four treatment groups were used in a 40-day chronic relapsing model of colitis: control diet, control diet with DSS added to drinking water, control diet adjusted with 10% by weight steamed broccoli sprouts, and 10% broccoli sprout diet with DSS added to drinking water (Figure 1).Bacterial communities were sampled from several locations in the gastrointestinal tract at the end of the study.

Figure S5
. Abundance of bacterial SVs identified as differential for the Broccoli+DSS group.Important features (SVs) were identified through permutational random forest analysis, and only the features important to this group (>50 reads) are listed out of 188 significant (p < 0.05) features across all treatments.Model accuracy was 98%.Bacterial sequence variants (SV) are identified as the lowest level of taxonomic identity possible, with "NA" indicating which could not be identified to species, and the number indicating which specific SV it was.Four treatment groups were used in a 34-day chronic relapsing model of colitis: control diet, control diet with DSS added to drinking water, control diet adjusted with 10% by weight steamed broccoli sprouts, and 10% broccoli sprout diet with DSS added to drinking water (Figure 1).Bacterial communities were sampled from several locations in the gastrointestinal tract at the end of the study.A total of 142 SVs were identified, and 95 had a proportion >1% and are visualized here.Four treatment groups were used in a 34-day chronic relapsing model of colitis: control diet, control diet with DSS added to drinking water, control diet adjusted with 10% by weight steamed broccoli sprouts, and 10% broccoli sprout diet with DSS added to drinking water (Figure 1).Bacterial communities were sampled from several locations in the gastrointestinal tract at the end of the study.Table S1 Targets  Gene block TGTTGAATCTGCCGGGCTGAAGGTTTTATCCTCACATTGCACAAGAGGATTGTCGA AAGAAGAATTAGCTTCCGGTGATTTTTCAAGTTCACTTCAATGGTGGGACCAGTGT ATTGCTGATCATA Protocol 1) 1 cycle at 50°C for 2 min; 2) 1 cycle at 95°C for 1 min; 3) 40 cycles at 95°C for 15 s, 60°C

Supplemental Tables
for 30 s and 72°C for 25 s, followed by a plate read,

Figure S2 .
Figure S2.Abundance of bacterial sequence variants identified as differential for the Control group.Important features (SVs) were identified through permutational random forest analysis, and only the features important to this group (>50 reads) are listed out of 188 significant (p < 0.05) features across all treatments.Model accuracy was 98%.Bacterial sequence variants (SV) are identified as the lowest level of taxonomic identity possible, with "NA" indicating which could not be identified to species, and the number indicating which specific SV it was.Four treatment groups were used in a 34-day chronic relapsing model of colitis: control diet, control diet with DSS added to drinking water, control diet adjusted with 10% by weight steamed broccoli sprouts, and 10% broccoli sprout diet with DSS added to drinking water (Figure1).Bacterial communities were sampled from several locations in the gastrointestinal tract at the end of the study.

Figure S6 .
Figure S6.Core bacterial sequence variants (SVs) shared by 70% of gut samples from A) mice consuming a 10% steamed broccoli sprout diet with or without DSS added to drinking water, or B) mice consuming a control diet with DSS in drinking water.There were no bacterial SVs shared across 70% of Control+DSS and Broccoli+DSS samples (data not shown).Four treatment groups were used in a 34-day chronic relapsing model of colitis: control diet, control diet with DSS added to drinking water, control diet adjusted with 10% by weight steamed broccoli sprouts, and 10% broccoli sprout diet with DSS added to drinking water (Figure1).Bacterial communities were sampled from several locations in the gastrointestinal tract at the end of the study.

Figure S7 :
Figure S7: qPCR results for the Bacterial operon BT2159-BT2156 for Bacteroides thetaiotaomicron (VPI-5482) in mice with and without chronic, replacing DSS-induced colitis where diet was manipulated.Note, mean copy number scales vary between charts.

Figure S8 .
Figure S8.Bacterial sequence variants which are hypothesized to be sourced from the cecum and creating population sinks in other locations in the intestines of mice.Potential source SVs were identified with the SourceTracker algorithm modified for the R platform.A total of 142 SVs were identified, and 95 had a proportion >1% and are visualized here.Four treatment groups were used in a 34-day chronic relapsing model of colitis: control diet, control diet with DSS added to drinking water, control diet adjusted with 10% by weight steamed broccoli sprouts, and 10% broccoli sprout diet with DSS added to drinking water (Figure1).Bacterial communities were sampled from several locations in the gastrointestinal tract at the end of the study.