Diet rapidly and reproducibly alters the human gutmicrobiome
Abstract
Long-term diet influences the structure and activity of the trillions ofmicroorganisms residing in the human gut1–5, but itremains unclear how rapidly and reproducibly the human gut microbiome respondsto short-term macronutrient change. Here, we show that the short-termconsumption of diets composed entirely of animal or plant products altersmicrobial community structure and overwhelms inter-individual differences inmicrobial gene expression. The animal-based diet increased the abundance ofbile-tolerant microorganisms (Alistipes, Bilophila, andBacteroides) and decreased the levels of Firmicutes thatmetabolize dietary plant polysaccharides (Roseburia, Eubacteriumrectale, and Ruminococcus bromii). Microbialactivity mirrored differences between herbivorous and carnivorousmammals2, reflectingtrade-offs between carbohydrate and protein fermentation. Foodborne microbesfrom both diets transiently colonized the gut, including bacteria, fungi, andeven viruses. Finally, increases in the abundance and activity ofBilophila wadsworthia on the animal-based diet support alink between dietary fat, bile acids, and the outgrowth of microorganismscapable of triggering inflammatory bowel disease6. In concert, these results demonstrate that thegut microbiome can rapidly respond to altered diet, potentially facilitating thediversity of human dietary lifestyles.
There is growing concern that recent lifestyle innovations, most notably thehigh-fat/high-sugar “Western” diet, have altered the genetic compositionand metabolic activity of our resident microorganisms (the human gutmicrobiome)7. Such diet-inducedchanges to gut-associated microbial communities are now suspected of contributing togrowing epidemics of chronic illness in the developed world, including obesity4,8and inflammatory bowel disease6. Yet,it remains unclear how quickly and reproducibly gut bacteria respond to dietary change.Work in inbred mice shows that shifting dietary macronutrients can broadly andconsistently alter the gut microbiome within a single day7,9. By contrast,dietary interventions in human cohorts have only measured community changes ontimescales of weeks10 tomonths4, failed to findsignificant diet-specific effects1, ordemonstrated responses among a limited number of bacterial taxa3,5.
Here, we examined if dietary interventions in humans can alter gut microbialcommunities in a rapid, diet-specific manner. We prepared two diets that variedaccording to their primary food source: a “plant-based diet”, which wasrich in grains, legumes, fruits, and vegetables; and an “animal-baseddiet”, which was composed of meats, eggs, and cheeses (Supplementary Table 1). We pickedthese sources to span the global diversity of modern human diets, which includesexclusively plant-based and nearly exclusively animal-based regimes11 (the latter being the case among somehigh-latitude and pastoralist cultures). Each diet was consumed adlibitum for five consecutive days by six male and four female Americanvolunteers between the ages of 21–33, whose body mass indices ranged from19–32 kg/m2 (Supplementary Table 2). Study volunteers were observed for four days beforeeach diet arm to measure normal eating habits (the baseline period) and for six daysafter each diet arm to assess microbial recovery (the washout period; Extended Data Fig. 1).Subjects’ baseline nutritional intake correlated well with their estimatedlong-term diet (Supplementary Table3). Our study cohort included a lifetime vegetarian (seeSupplementary Discussion, Extended Data Fig. 2, and Supplementary Table 4 for a detailed analysis of his diet and gutmicrobiota).
Each diet arm significantly shifted subjects' macronutrient intake (Fig. 1a–c). On the animal-based diet, dietaryfat increased from 32.5±2.2% to 69.5±0.4% kcal anddietary protein increased from 16.2±1.3% to 30.1±0.5%kcal (p<0.01 for both comparisons, Wilcoxon signed-rank test; Supplementary Table 5). Fiberintake was nearly zero, in contrast to baseline levels of 9.3±2.1 g/1,000kcal.On the plant-based diet, fiber intake rose to 25.6±1.1 g/1,000kcal, while bothfat and protein intake declined to 22.1±1.7% and10.0±0.3%, respectively (p<0.05 for all comparisons).Subjects’ weights on the plant-based diet remained stable, but decreasedsignificantly by day 3 of the animal-based diet (q<0.05, Bonferroni-correctedMann-Whitney U test; Extended DataFig. 3). Differential weight loss between the two diets cannot be explainedsimply by energy intake, as subjects consumed equal numbers of calories on the plant-and animal-based diets (1,695±172 kcal and 1,777±221 kcal, respectively;p=0.44).
To characterize temporal patterns of microbial community structure, we performed16S rRNA gene sequencing on samples collected each day of the study (Supplementary Table 6). Wequantified the microbial diversity within each subject at a given time-point(α-diversity) and the difference between each subjects' baseline anddiet-associated gut microbiota (β-diversity) (Fig.1d,e). Although no significant differences in α-diversity weredetected on either diet, we observed a significant increase in β-diversity thatwas unique to the animal-based diet (q<0.05, Bonferroni-corrected Mann-Whitney Utest). This change occurred a single day after the diet reached the distal gutmicrobiota (as indicated by the food tracking dye; Extended Data Fig. 3a).Subjects’ gut microbiota reverted to their original structure 2 days after theanimal-based diet ended (Fig. 1e).
Analysis of the relative abundance of bacterial taxonomic groups supported ourfinding that the animal-based diet had a greater impact on the gut microbiota than theplant-based diet (Fig. 2). We hierarchicallyclustered species-level bacterial phylotypes by the similarity of their dynamics acrossdiets and subjects (see SupplementaryMethods and SupplementaryTables 7 and 8). Statistical testing identified 22 clusters whose abundancesignificantly changed while on the animal-based diet, while only 3 clusters showedsignificant abundance changes while on the plant-based diet (q<0.05, Wilcoxonsigned-rank test; Supplementary Table9). Notably, the genus Prevotella, one of the leadingsources of inter-individual gut microbiota variation12 and hypothesized to be sensitive to long-term fiberintake1,13, was reduced in our vegetarian subject duringconsumption of the animal-based diet (see Supplementary Discussion). Wealso observed a significant positive correlation between subjects’ fiber intakeover the past year and baseline gut Prevotella levels (Extended Data Fig. 4 and Supplementary Table 10).
To identify functional traits linking clusters that thrived on the animal-baseddiet, we selected the most abundant taxon in the three most-enriched clusters(Bilophila wadsworthia, Cluster 28; Alistipesputredinis, Cluster 26; and a Bacteroides sp., Cluster29), and performed a literature search for their lifestyle traits. That search quicklyyielded a common theme of bile-resistance for these taxa, which is consistent withobservations that high fat intake causes more bile acids to be secreted14.
Analysis of fecal SCFAs and bacterial clusters suggests that macronutrient shiftson both diets also altered microbial metabolic activity. Relative to the plant-baseddiet and baseline samples, the animal-based diet resulted in significantly lower levelsof the products of carbohydrate fermentation and a higher concentration of the productsof amino acid fermentation (Fig. 3a,b; Supplementary Table 11). When wecorrelated subjects’ SCFA concentrations with the same-day abundance ofbacterial clusters from Fig. 2, we foundsignificant positive relationships between clusters composed of putrefactivemicrobes15,16 (i.e. Alistipesputredinis and Bacteroides spp.) and SCFAs that are theend products of amino acid fermentation (Extended Data Fig. 5). We also observed significant positivecorrelations between clusters comprised of saccharolytic microbes3 (e.g.Roseburia, E. rectale, and F.prausnitzii) and the products of carbohydrate fermentation.
In order to test whether the observed changes in microbial community structureand metabolic end products were accompanied by more widespread shifts in the gutmicrobiome, we measured microbial gene expression using RNA sequencing (RNA-Seq). Asubset of samples was analyzed, targeting the baseline periods and the final 2 days ofeach diet (Extended Data Fig.1, Supplementary Table12). We identified several differentially-expressed metabolic modules andpathways during the plant- and animal-based diets (Supplementary Tables 13 and 14).The animal-based diet was associated with increased expression of key genes for vitaminbiosynthesis (Fig. 3c); the degradation ofpolycyclic aromatic hydrocarbons (Fig. 3d), whichare carcinogenic compounds produced during the charring of meat17; and the increased expression ofβ-lactamase genes (Fig. 3e). Metagenomicmodels constructed from our 16S rRNA data18 suggest that the observed expression differences are due to acombination of regulatory and taxonomic shifts within the microbiome (Supplementary Tables 15 and16).
We next hierarchically-clustered microbiome samples based on the transcriptionof KEGG orthologous groups19, whichsuggested that overall microbial gene expression was strongly linked to host diet.Nearly all of the diet samples could be clustered by diet arm (p<0.003,Fisher’s exact test; Fig. 3f), despite thepre-existing inter-individual variation we observed during the baseline diets (Extended Data Fig. 6a,b). Still,subjects maintained their inter-individual differences on a taxonomic level on the dietarms (Extended Data Fig. 6c).Of the three RNA-Seq samples on the animal-based diet that clustered with samples fromthe plant-based diet, all were taken on day 3 of the diet arm. In contrast, all RNA-Seqsamples from the final day of the diet arms (day 4) clustered by diet (Fig. 3f).
Remarkably, the plant- and animal-based diets also elicited transcriptionalresponses that were consistent with known differences in gene abundance between the gutmicrobiomes of herbivorous and carnivorous mammals, such as the tradeoffs between aminoacid catabolism versus biosynthesis, and in the interconversions of phosphoenolpyruvate(PEP) and oxaloacetate2 (Fig. 3g,h). The former pathway favors amino acidcatabolism when protein is abundant2,and we speculate that the latter pathway produces PEP for aromatic amino acid synthesiswhen protein is scarce20. In all 14steps of these pathways, we observed fold-changes in gene expression on the plant- andanimal-based diets whose directions agreed with the previously reported differencesbetween herbivores and carnivores (p<0.001, Binomial test). Notably, thisperfect agreement is not observed when the plant- and animal-based diets are onlycompared to their respective baseline periods, indicating that the expression patternsin Fig. 3g,h reflect functional changes from bothdiet arms (Supplementary Table17).
Our findings that the human gut microbiome can rapidly switch betweenherbivorous and carnivorous functional profiles may reflect past selective pressuresduring human evolution. Consumption of animal foods by our ancestors was likelyvolatile, depending on season and stochastic foraging success, with readily availableplant foods offering a fallback source of calories and nutrients21. Microbial communities that couldquickly, and appropriately, shift their functional repertoire in response to diet changewould have subsequently enhanced human dietary flexibility. Examples of this flexibilitymay persist today in the form of the wide diversity of modern human diets11.
We next examined if, in addition to affecting the resident gut microbiota, eitherdiet arm introduced foreign microorganisms into the distal gut. We identified foodbornebacteria on both diets using 16S rRNA gene sequencing. The cheese and cured meatsincluded in the animal-based diet were dominated by lactic acid bacteria commonly usedas starter cultures for fermented foods22,23: Lactococcuslactis, Pediococcus acidilactici, andStreptococcus thermophilus (Fig.4a). Common non-lactic acid bacteria included severalStaphylococcus taxa; strains from this genus are often used whenmaking fermented sausages23. Duringthe animal-based diet, three of the bacteria associated with cheese and cured meats(L. lactis, P. acidilactici, andStaphylococcus) became significantly more prevalent in fecalsamples (p<0.05, Wilcoxon signed-rank test; Extended Data Fig. 7c),indicating that bacteria found in common fermented foods can reach the gut at abundancesabove the detection limit of our sequencing experiments (on average 1 in4×104 gut bacteria; Supplementary Table 6).
We also sequenced the internal transcribed spacer (ITS) region of the rRNAoperon from community DNA extracted from food and fecal samples to study therelationship between diet and enteric fungi, which to date remains poorly characterized(Supplementary Table 18).Menu items on both diets were colonized by the genera Candida,Debaryomyces, Penicillium, andScopulariopsis (Fig. 4a andExtended Data Fig. 7a),which are often found in fermented foods22. A Penicillium sp. and Candidasp. were consumed in sufficient quantities on the animal- and plant-based diets to showsignificant ITS sequence increases on those respective diet arms (Extended Data Fig. 7b,c).
Microbial culturing and re-analysis of our RNA-Seq data suggested that foodbornemicrobes survived transit through the digestive system and may have been metabolicallyactive in the gut. Mapping RNA-Seq reads to an expanded reference set of 4,688 genomes(see Supplementary Methods)revealed a significant increase on the animal-based diet for transcripts expressed byfood-associated bacteria (Fig. 4b–d) andfungi (Fig. 4e; q<0.1, Kruskal-Wallistest). Many dairy-associated microbes remained viable after passing through thedigestive tract, as we isolated 19 bacterial and fungal strains with high geneticsimilarity (>97% ITS or 16S rRNA) to microbes cultured from cheeses fedto the subjects (Supplementary Table19). Moreover, L. lactis was more abundant in fecal culturessampled after the animal-based diet, relative to samples from the preceding baselineperiod (p<0.1; Wilcoxon Signed-Rank test). We also detected an overall increasein the fecal concentration of viable fungi on the animal-based diet (Fig. 4f; p<0.02; Mann-Whitney U test).Interestingly, we detected RNA transcripts from multiple plant viruses Extended Data Fig. 8). One plantpathogen, Rubus chlorotic mottle virus, was only detectable on the plant-based diet(Fig. 4g). This virus infects spinach24, which was a key ingredient in theprepared meals on the plant-based diet. These data support the hypothesis that plantpathogens can reach the human gut via consumed plant matter25.
Finally, we found that microbiota changes on the animal-based diet could belinked to altered fecal bile acid profiles and the potential for human enteric disease.Recent mouse experiments have shown high-fat diets lead to increased enteric deoxycholicconcentrations (DCA); this secondary bile acid is the product of microbial metabolismand promotes liver cancer26. In ourstudy, the animal-based diet significantly increased the levels of fecal DCA (Fig. 5a). Expression of bacterial genes encodingmicrobial bile salt hydrolases, which are prerequisites for gut microbial production ofDCA27, also exhibitedsignificantly higher expression on the animal-based diet (Fig. 5b). Elevated DCA levels in turn, may have contributed to the microbialdisturbances on the animal-based diet, as this bile acid can inhibit the growth ofmembers of the Bacteroidetes and Firmicutes phyla28.
Mouse models have also found evidence that inflammatory bowel disease can becaused by B. wadsworthia, a sulfite-reducing bacterium whose productionof H2S is thought to inflame intestinal tissue6. Growth of B. wadsworthia is stimulatedin mice by select bile acids secreted while consuming saturated fats from milk. Ourstudy provides several lines of evidence confirming that B. wadsworthiagrowth in humans can also be promoted by a high-fat diet. First, we observed B.wadsworthia to be a major component of the bacterial cluster that increasedmost strongly while on the animal-based diet (C28; Fig.2 and Supplementary Table8). This Bilophila-containing cluster also showedsignificant positive correlations with both long-term dairy (p<0.05; Spearmancorrelation) and baseline saturated fat intake (Supplementary Table 20), supporting the proposed link tomilk-associated saturated fats6.Second, the animal-based diet led to significantly increased fecal bile acidconcentrations (Fig. 5c and Extended Data Fig. 9). Third, weobserved significant increases in the abundance of microbial DNA and RNA encodingsulfite reductases on the animal-based diet (Fig.5d,e). Together, these findings are consistent with the hypothesis thatdiet-induced changes to the gut microbiota may contribute to the development ofinflammatory bowel disease. More broadly, our results emphasize that a morecomprehensive understanding of diet-related diseases will benefit from elucidating linksbetween nutritional, biliary, and microbial dynamics.
Methods
Sample collection
We recruited 11 unrelated subjects (n=10 per diet; 9 individualscompleted both arms of the study). One participant suffered from a chronicgastrointestinal disease, but all other volunteers were otherwise healthy. Thevolunteers’ normal bowel frequencies ranged from three times a day toonce every other day. Three participants had taken antibiotics in the past year.Additional subject information is provided in Supplementary Table 2.Gut microbial communities were sampled from feces. Subjects were instructed tocollect no more than one sample per day, but to log all bowel movements. Nomicrobiota patterns were observed as a function of sampling time of day (datanot shown). Subjects collected samples by placing disposable commode specimencontainers (Claflin Medical Equipment, Warwick, RI) under their toilet seatsbefore bowel movements. CultureSwabs™ (BD, Franklin Lakes, NJ) were thenused to collect fecal specimens for sequencing analyses, and larger collectiontubes were provided for harvesting larger, intact stool samples (~10g)for metabolic analyses. Each sample was either frozen immediately at−80°C or briefly stored in personal −20°Cfreezers before transport to the laboratory.
Diet design
We constructed two diet arms, each of which consisted mostly of plant-or animal-based foods (Extended Data Fig. 1). Subjects on the plant-based diet ate cerealfor breakfast and precooked meals made of vegetables, rice, and lentils forlunch and dinner (see Supplementary Table 1 for a full list of diet ingredients). Freshand dried fruits were provided as snacks on this diet. Subjects on theanimal-based diet ate eggs and bacon for breakfast, and cooked pork and beef forlunch. Dinner consisted of cured meats and a selection of four cheeses. Snackson this diet included pork rinds, cheese, and salami. Ingredients for theplant-based diet, dinner meats and cheeses for the animal-based diet, and snacksfor both diets were purchased from grocery stores. Lunchmeats for theanimal-based diet were prepared by a restaurant that was instructed to not addsauce to the food. On each diet arm, subjects were instructed to eat onlyprovided foods or allowable beverages (water or unsweetened tea for both diets;coffee was allowed on the animal-based diet). They were also allowed to add 1salt packet per meal, if desired for taste. Subjects could eat unlimited amountsof the provided foods. Outside of the five-day diet arms, subjects wereinstructed to eat normally.
Food logs, subject metadata, and dietary questionnaires
Subjects were given notepads to log their diet, health, and bowelmovements during the study. Subjects transcribed their notepads into digitalspreadsheets when the study ended. Each ingested food (including foods on thediet arm) was recorded, as well as data on time, location, portion size, andfood brand. Subjects were provided with pocket digital scales (American Weigh,Norcross, GA) and a visual serving size guide to aid with quantifying the amountof food consumed. Each day, subjects tracked their weight using either a scaleprovided in the lab, or their own personal scales at home. While on theanimal-based diet, subjects were requested to measure their urinary ketonelevels using provided Ketostix strips (Bayer, Leverkusen, Germany; Extended Data Fig. 1). ifsubjects recorded a range of ketone levels (the Ketostix color key uses arange-based reporting system), the middle value of that range was used forfurther analysis. Subjects were encouraged to record any discomfort theyexperienced while on either diet (e.g. bloating, constipation).Subjects tracked all bowel movements, regardless of whether or not theycollected samples, recording movement time, date, and location, andqualitatively documented stool color, odor, and type1. Subjects were also asked to report when they observedstool staining from food dyes consumed at the beginning and end of each diet arm(Extended Data Fig.3a).
Diet quantification
We quantified subjects' daily nutritional intake during thestudy using CalorieKing and Nutrition Data System for Research (NDSR). TheCalorieKing™ food database (La Mesa, CA) was accessed via theCalorieKing Nutrition & Exercise Manager software (version 4.1.0).Subjects' food items were manually transferred from digital spreadsheetsinto the CalorieKing software, which then tabulated each food'snutritional content. Macronutrient content per serving was calculated for eachof the prepared meals on the animal- and plant-based diet using lists of thosemeals’ ingredients. Nutritional data was outputted from CalorieKing inCSV format and parsed for further analysis using a custom Python script. NDSRintake data were collected and analyzed using Nutrition Data System for Researchsoftware version 2012, developed by the Nutrition Coordinating Center (NCC),University of Minnesota, Minneapolis, MN. We estimated subjects'long-term diet using the National Cancer Institute’s Diet HistoryQuestionnaire II2 (DHQ). We used the DHQto quantify subjects' annual diet intake, decomposed into 176nutritional categories. Subjects completed the yearly, serving size-includedversion of the DHQ online using their personal computers. We parsed thesurvey's results using the Diet*Calc software (version 1.5; Risk FactorMonitoring and Methods Branch, NCI) and its supplied 'Food and NutrientDatabase', and 'dhqweb.yearly.withserv.2010.qdd' QDDfile.
There was good agreement between subjects’ diets as measured byCalorieKing, the NDSR, and the DHQ: 18 of 20 nutritional comparisons betweenpairs of databases showed significant correlations (Supplementary Table 3).Unless specified, nutritional data presented in this manuscript reflectCalorieKing measurements.
16S rRNA gene sequencing and processing
Temporal patterns of microbial community structure were analyzed fromdaily fecal samples collected across each diet (Extended Data Fig. 1).Samples were kept at −80°C until DNA extraction with thePowerSoil bacterial DNA extraction kit (MoBio, Carlsbad CA). The V4 region ofthe 16S rRNA gene was PCR amplified in triplicate, and the resulting ampliconswere cleaned, quantified, and sequenced on the Illumina HiSeq platform accordingto published protocols3,4 and using custom barcoded primers (Supplementary Table 6).Raw sequences were processed using the QIIME software package (QuantitativeInsights Into Microbial Ecology)5. Only full-length, high-quality reads (−r=0) were used for analysis. Operational taxonomic units (OTUs) were pickedat 97% similarity against the Greengenes database6 (constructed by thenested_gg_workflow.py QiimeUtils script on 4 Feb 2011), which we trimmed to spanonly the 16S rRNA region flanked by our sequencing primers (positions521–773). In total, we characterized an average of 43,589±1,82616S rRNA sequences for 235 samples (an average of 0.78 samples per person perstudy day; Supplementary Table6). Most of the subsequent analysis of 16S rRNA data, includingcalculations of α- and β-diversity, were performed using customPython scripts, the SciPy Python library7, and the Pandas Data Analysis Library8. Correction for multiple hypothesis testing utilized the fdrtool9R library, except in the case of small test numbers, in which case theBonferroni correction was used.
OTU clustering
We used clustering to simplify the dynamics of thousands of OTUs into alimited number of variables that could be more easily visualized and manuallyinspected. Clustering was performed on normalized OTU abundances. Suchabundances are traditionally computed by scaling each sample's reads tosum to a fixed value (e.g. unity); this technique is intendedto account for varying sequencing depth between samples. However, this standardtechnique may cause false relationships to be inferred between microbial taxa,as increases in the abundance of one microbial group will cause decreases in thefractional abundance of other microbes (this artifact is known as a“compositional” effect10). To avoid compositional biases, we employed an alternativenormalization approach, which instead assumes that no more than half of the OTUsheld in common between two microbial samples change in abundance. This methoduses a robust (outlier-resistant) regression to estimate the median OTUfold-change between communities, by which it subsequently rescales all OTUs.
To further simplify community dynamics, we only included in ourclustering model OTUs that comprised 95% of total reads (after rankingby normalized abundance). Abundances for each included OTU were then convertedto log-space and mediancentered.
We computed OTU pairwise distances using the Pearson correlation (OTUabundances across all subjects and time points were used). The resultingdistance matrix was subsequently input into Scipy's hierarchicalclustering function ('fcluster'). Default parameters were usedfor fcluster, with the exception of the clustering criterion, which was set to'distance', and the clustering threshold, which was set to'0.7'. These parameters were selected manually so that clusterboundaries visually agreed with the correlation patterns plotted in a matrix ofpairwise OTU distances.
Statistics on cluster abundance during baseline and diet periods werecomputed by taking median values across date ranges. Baseline date ranges werethe 4 days preceding each diet arm (i.e. days −4 through −1).Date ranges for the diet arms were chosen so as to capture the full effects ofeach diet. These ranges were not expected to perfectly overlap with the dietarms themselves, due to the effects of diet transit time. We therefore chosediet arm date ranges that accounted for transit time (as measured by fooddye; Extended Data Fig.3a), picking ranges that began 1 day after foods reached the gut, andended 1 day before the last diet arm meal reached the gut. These criteria ledmicrobial abundance measurements on the plant-based diet to span days2–4 of that study arm, and animal-based diet measurements to span days2–5 of that diet arm.
RNA-Seq sample preparation and sequencing
In order to test if the observed changes in community structure wereaccompanied by changes to the active subset of the human gut microbiome, wemeasured communitywide gene expression using meta-transcriptomics11–14 (RNA sequencing, RNA-Seq; Supplementary Table 12). Samples were selected based on ourprior 16S rRNA gene sequencing-based analysis, representing 3 baseline days and2 timepoints on each diet (n=5–10 samples/timepoint; Extended Data Fig. 1).Microbial cells were lysed by a bead beater (BioSpec Products, Bartlesville,OK), total RNA was extracted with phenol:chloroform:isoamyl alcohol (pH 4.5,125:24:1, Ambion 9720) and purified using Ambion MEGAClear columns (LifeTechnologies, Grand Island, NY), and rRNA was depleted via Ambion MICROBExpresssubtractive hybridization (Life Technologies, Grand Island, NY) and customdepletion oligos. The presence of genomic DNA contamination was assessed by PCRwith universal 16S rRNA gene primers. cDNA was synthesized using SuperScript IIand random hexamers ordered from Invitrogen (Life Technologies, Grand Island,NY), followed by second strand synthesis with RNaseH and E.coliDNA polymerase (New England Biolabs, Ipswich, MA). Samples were prepared forsequencing with an Illumina HiSeq instrument after enzymatic fragmentation(NEBE6040L/M0348S). Libraries were quantified by quantitative reversetranscriptase PCR (qRT-PCR) according to the Illumina protocol. qRT-PCR assayswere run using ABsoluteTM QPCR SYBR® Green ROX Mix (Thermo Scientific,Waltham, MA) on a Mx3000P QPCR System instrument (Stratagene, La Jolla, CA). Thesize distribution of each library was quantified on an Agilent HS-DNA chip.Libraries were sequenced using the Illumina HiSeq platform.
Functional analysis of RNA-Seq data
We used a custom reference database of bacterial genomes to performfunctional analysis of the RNA-Seq data12. This reference included 538 draft and finished bacterial genomesobtained from human-associated microbial isolates15, and the Eggerthella lenta DSM2243 referencegenome. All predicted proteins from the reference genome database were annotatedwith KEGG16 orthologous groups (KOs)using the KEGG database (version 52; BLASTX e-value<10–5, Bitscore>50, and >50% identity). for query genes withmultiple matches, the annotated reference gene with the lowest evalue was used.When multiple annotated genes with an identical e-value were encountered after aBLAST query, we included all KOs assigned to those genes. Genes from thedatabase with significant homology (BLASTN e-value<10–20) tonon-coding transcripts from the 539 microbial genomes were excluded fromsubsequent analysis. High-quality reads (see Supplementary Table 12for sequencing statistics) were mapped using SSAHA217, to our reference bacterial database and the Illuminaadaptor sequences (SSAHA2 parameters: “-best 1 -score 20-solexa”). The number of transcripts assigned to each gene was thentallied and normalized to reads per kilobase per million mapped reads (RPKM). Toaccount for genes that were not detected due to limited sequencing depth, apseudocount of 0.01 was added to all samples. Samples were clustered in Matlab(version 7.10.0) using a Spearman distance matrix (commands: pdist, linkage, anddendrogram). Genes were grouped by taxa, genomes, and KEGG orthologous groups(KOs) by calculating the cumulative RPKM for each sample. HUMAnN18 was used for metabolic reconstructionfrom metagenomic data followed by LefSe19analysis to identify significant biomarkers. A modified version of the“SConstruct” file was used to input KEGG orthologous groupcounts into the HUMAnN pipeline for each RNA-Seq dataset. We then ran LefSe onthe resulting KEGG module abundance file using the “-o 1000000”flag.
Taxonomic analysis of RNA-Seq data
We used Bowtie 2 read alignment program20 and the Integrated Microbial Genomes (IMG; version 3.5)database21 to map RNA-Seq reads to acomprehensive reference survey of prokaryotic, eukaryotic, and viral genomes.Our reference survey included all 2,809 viral genomes in IMG (as of version3.5), a set of 1,813 bacterial and archaeal genomes selected to minimize strainredundancy22, and 66 genomes spanningthe Eukarya except for the plants and non-nematode Bilateria. Reads were mappedto reference genomes using Bowtie, which was configured to analyze matedpaired-end reads, and return fragments with a minimum length of 150bp and amaximum length of 600bp. All other parameters were left to their default values.The number of base pairs in the reference genome dataset exceededBowtie's reference size limit, so we split the reference genomes intofour subsets. Each read was mapped to each of these four subreference datasets,and the results were merged by picking the highest-scoring match across thesub-references. We settled tied scores by randomly choosing one of thebest-scoring matches. To more precisely measure the presence or absence ofspecific taxa, we next filtered out reads that mapped to more than referencesequence. Raw read counts were computed for each reference genome by countingthe number of reads that mapped to coding sequences according to the IMGannotations; these counts were subsequently normalized using RPKM scaling. Ouranalysis pipeline associated several sequences with marine algae, which areunlikely to colonize the human gut. We also detected a fungal pathogenexclusively in samples from subjects consuming the animal-based diet(Neosartorya fischeri); this taxon was suspected of being amisidentified cheese fungus, due to its relatedness toPenicillium. We thus reanalysed protist and N.fischeri reads associated with potentially mis-annotated taxa usingBLAST searches against the NCBI non-redundant database, and we assigned taxonomymanually based on the most common resulting hits (Extended Data Fig.8).
Quantitative PCR
Community DNA was isolated with the PowerSoil bacterial DNA extractionkit (MoBio, Carlsbad CA). To determine the presence of hydrogen consumers, PCRwas performed on fecal DNA using the following primer sets: (i) Sulfitereductase23 (dsrA), F-5’-CCAACATGCACGGYT CCA-3’,R-5’-CGTCGAACTTGAACTTGAACTTGTAGG-3’; and (ii) Sulfatereduction24,25 (aps reductase),F-5’-TGGCAGATMATGATYMACGG-3’, R-5’-GGGCCGTAACCGTCCTTGAA-3’. qPCR assays were run using ABsoluteTM QPCRSYBR® Green ROX Mix (Thermo Scientific, Waltham, MA) on a Mx3000P QPCRSystem instrument (Stratagene, La Jolla, CA). Fold-changes were calculatedrelative to the 16S rRNA gene using the 2-ΔΔCt method and thesame primers used for 16S rRNA gene sequencing.
Short-chain fatty acid measurements
Fecal SCFA content was determined by gas chromatography. Chromatographicanalysis was carried out using a Shimadzu GC14-A system with a flame ionizationdetector (FID) (Shimadzu Corp, Kyoto, Japan). Fused silica capillary columns 30m× 0.25 mm coated with 0.25um film thickness were used (Nukol™for the volatile acids and SPB™-1000 for the nonvolatile acids (SupelcoAnalytical, Bellefonte, PA). Nitrogen was used as the carrier gas. The oventemperature was 170°C and the FID and injection port was set to225°C. The injected sample volume was 2 µL and the run time foreach analysis was 10 minutes. The chromatograms and data integration was carriedout using a Shimadzu C-R5A Chromatopac. A volatile acid mix containing 10 mM ofacetic, propionic, isobutyric, butyric, isovaleric, valeric, isocaproic,caproic, and heptanoic acids was used (Matreya, Pleasant Gap, PA). Anon-volatile acid mix containing 10 mM of pyruvic and lactic and 5 mM ofoxalacetic, oxalic, methy malonic, malonic, fumaric, and succinic was used(Matreya, Pleasant Gap, PA). A standard stock solution containing 1%2-methyl pentanoic acid (Sigma-Aldrich, St. Louis, MO) was prepared as aninternal standard control for the volatile acid extractions. A standard stocksolution containing 50 mM benzoic acid (Sigma-Aldrich, St. Louis, MO) wasprepared as an internal standard control for the non-volatile acidextractions.
Samples were kept frozen at −80°C until analysis. Thesamples were removed from the freezer and 1,200µL of water was added toeach thawed sample. The samples were vortexed for 1 minute until the materialwas homogenized. The pH of the suspension was adjusted to 2–3 by adding50 µL of 50% sulfuric acid. The acidified samples were kept atroom temperature for 5 minutes and vortexed briefly every minute. The sampleswere centrifuged for 10 minutes at 5,000g. 500 µL of the clearsupernatant was transferred into two tubes for further processing. For thevolatile extraction 50 µL of the internal standard (1% 2-methylpentanoic acid solution) and 500 µL of ethyl ether anhydrous were added.The tubes were vortexed for 30 seconds and then centrifuged at 5,000g for 10minutes. 1 µL of the upper ether layer was injected into thechromatogram for analysis. For the nonvolatile extraction 50 µL of theinternal standard (50 mM benzoic acid solution) and 500 µL of borontrifluoride-methanol solution (Sigma-Aldrich St. Louis, MO) were added to eachtube. These tubes were incubated overnight at room temperature. 1 mL of waterand 500 µL of chloroform were added to each tube. The tubes werevortexed for 30 seconds and then centrifuged at 5,000g for 10 minutes. 1µL of the lower chloroform layer was injected into the chromatogram foranalysis. 500 µL of each standard mix was used and the extracts preparedas described for the samples. The retention times and peak heights of the acidsin the standard mix were used as references for the sample unknowns. These acidswere identified by their specific retention times and the concentrationsdetermined and expressed as mM concentrations per gram of sample.
Bulk bile acid quantification
Fecal bile acid concentration was measured as described previously26. 100 mg of lyophilized stool was heatedto 195°C in 1 mL of ethylene glycol KOH for 2 hours, neutralized with 1mL of saline and 0.2 mL of concentrated HCl, and extracted into 6 mL of diethylether 3 times. After evaporation of the ether, the sample residues weredissolved in 6 mL of methanol and subjected to enzymatic analysis. Enzymaticreaction mixtures consisted of 66.5 mmol/L Tris, 0.33 mmol/L EDTA, 0.33 mol/Lhydrazine hydrate, 0.77 mmol/L NAD (N 7004, Sigma-Aldrich, St. Louis, MO),0.033U/mL 3〈- hydroxysteroid dehydrogenase (Sigma-Aldrich, St. Louis,MO) and either sample or standard (taurocholic acid; Sigma-Aldrich, St. Louis,MO) dissolved in methanol. After 90 minutes of incubation at 37°C,absorbance was measured at 340 nm.
Measurement of primary and secondary bile acids
Profiling of fecal primary and secondary bile acids was performed usinga modified version of a method described previously27. To a suspension of ~100 mg of stool and 0.25 mLof water in a 1 dram Teflon-capped glass vial was added 200 mg of glass beads.The suspension was homogenized by vortexing for 60–90 seconds. Ethanol(1.8 mL) was added, and the suspension was heated with stirring in a heatingblock at 80°C for 1.5 h. The sample was cooled, transferred to a 2 mLEppendorf tube, and centrifuged at 13500 rpm for 1–2 min. Thesupernatant was removed and retained. The pellet was resuspended in 1.8 mL of80% aqueous ethanol, transferred to the original vial, and heated to80°C for 1.5 h. The sample was centrifuged again, and the supernatantwas removed and added to the first extraction supernatant. The pellet wasresuspended in 1.8 mL of chloroform:methanol (1:1 v/v) and refluxed for30–60 min. The sample was centrifuged, and the supernatant removed andconcentrated to dryness on a rotary evaporator. The ethanolic supernatants wereadded to the same flask, the pH was adjusted to neutrality by adding aqueous0.01N HCl, and the combined extracts were evaporated to dryness. The driedextract was resuspended in 1 mL of 0.01N aqueous HCl by sonication for 30 min. ABIO-RAD Poly-Prep chromatography column (0.8×4cm) was loaded withLipidex 1000 as a slurry in MeOH, allowed to pack under gravity to a finalvolume of 1.1 mL, and washed with 10 mL of distilled water. The suspension wasfiltered through the bed of Lipidex 1000 and the effluent was discarded. Theflask was washed with 3 × 1 mL of 0.01N HCl, the washings were passedthrough the gel, and the bed was washed with 4 mL of distilled water. Bile acidsand sterols were recovered by elution of the Lipidex gel bed with 8 mL ofmethanol. A BIO-RAD Poly-Prep chromatography column (0.8×4cm) was loadedwith washed SP-Sephadex as a slurry in 72% aqueous MeOH to a finalvolume of 1.1 mL. The methanolic extract was passed through the SP-Sephadexcolumn, and the column was washed with 4 mL of 72% aqueous methanol. Theextract and wash were combined, and the pH was brought to neutral with 0.04Naqueous NaOH. A BIO-RAD Poly-Prep chromatography column (0.8×4cm) wasloaded with Lipidex-DEAP, prepared in the acetate form, as a slurry in72% aqueous MeOH to a final volume of 1.1 mL. The combined neutralizedeffluent was applied to the column, and the solution was eluted using air gaspressure (flow rate ~25 mL/h). The flask and column were washed with 2× 2 mL of 72% aqueous ethanol, and the sample and washings werecombined to give a fraction of neutral compounds including sterols. Unconjugatedbile acids were eluted using 4 mL of 0.1 M acetic acid in 72% (v/v)aqueous ethanol that had been adjusted to pH 4.0 by addition of concentratedammonium hydroxide. The fraction containing bile acids was concentrated ztodryness on a rotary evaporator.
The bile acids were converted to their corresponding methyl esterderivatives by the addition of 0.6 mL of MeOH followed by 40 µL of a 2.0M solution of (trimethylsilyl)diazomethane in diethyl ether. The solution wasdivided in half, and each half of the sample was concentrated to dryness on arotary evaporator. The bile acids in the first half of the sample were convertedto their corresponding trimethylsilyl ether derivatives by the addition of 35µL of a 2:1 solution of N,Obis(trimethylsilyl)trifluoroacetamide andchlorotrimethylsilane and analyzed by GC-MS. The identities of individual bileacids were determined by comparison of retention time and fragmentation patternto known standards. Both the ratio of cholest-3-ene to deoxycholic acid in thesample and the amount of internal standard to be added were determined byintegrating peak areas. A known amount of the internal standard, 5®-cholestane-3®-ol (5®-coprostanol), was added to the second halfof the sample (0.003– 0.07 mmol). The bile acids in the second half ofthe sample were converted to their corresponding trimethylsilyl etherderivatives by the addition of 35 µL of a 2:1 solution ofN,O-bis(trimethylsilyl)trifluoroacetamide and chlorotrimethylsilane and analyzedby GCMS. Amounts of individual bile acids were determined by dividing integratedbile acid peak area by the internal standard peak area, multiplying by theamount of internal standard added, and then dividing by half of the mass offecal matter extracted. In the event that the first half of the sample containedcholest-3-ene, the coprostanol peak area in the second half of the sample wascorrected by subtracting the area of the cholest-3-ene peak, determined byapplying the cholest-3-ene:deoxycholic acid ratio calculated from the first halfof the sample.
ITS sequencing
Fungal amplicon libraries were constructed with primers that target theinternal transcribed spacer (ITS), a region of the nuclear ribosomal RNA cistronshown to promote successful identification across a broad range of fungaltaxa28. We selected primers(ITS1f29 and ITS230) focused on the ITS1 region because itprovided the best discrimination between common cheese-associated fungi inpreliminary in silico tests. Multiplex capability was achievedby adding Golay barcodes to the ITS2 primer. Due to relatively lowconcentrations, fungal DNA was amplified in three serial PCR reactions, with thefirst reaction using 1 ul of the PowerSoil DNA extract, and the subsequent tworeactions using 1 ul of the preceding PCR product as the template. In each roundof PCR, sample reactions were performed in triplicate and then combined.Barcoded amplicons were cleaned, quantified and pooled to achieve approximatelyequal amounts of DNA from each sample using methods identical to those used for16S. We gel purified the pool, targeting amplicons between 150 bp and 500 bp insize, and submitted it for Illumina sequencing.
Preliminary taxonomic assignments of ITS reads using the 12_11 UNITEOTUs ITS database (see http://qiime.org) resulted in manyunassigned reads. To improve the percentage of reads assigned, we created ourown custom database of ITS1 sequences. We extracted ITS sequences from GenBankby targeting specific collections of reliable ITS sequences (e.g. AFTOL, FungalBarcoding Consortium) and by searching for sequences of yeasts and filamentousfungi that have been previously isolated from dairy and other food ecosystems.We also retrieved a wider range of fungi for our database by searching GenBankwith the query internal transcribed spacer[All Fields] AND fungi NOT‘uncultured’. Sequences that did not contain thefull ITS1 were removed. We also included reference OTUs that were identified aswidespread cheese fungi in a survey of cheese rinds (Wolfe, Button, and Dutton,unpublished data), but were not in public databases.
Microbial culturing
Fecal samples were cultured under conditions permissive for growth offood-derived microbes. Fecal samples were suspended in a volume ofphosphate-buffered saline (PBS) equivalent to ten times their weight. Serialdilutions were prepared and plated on brain heart infusion agar (BD Biosciences,San Jose, CA), supplemented with 100ug/ml cycloheximide, an antifungal agent,and plate count agar with milk and salt (per liter: 5g tryptone, 2.5g yeastextract, 1g dextrose, 1g whole milk powder, 30g NaCl, 15g agar) supplementedwith 50ug/ml chloramphenicol, an antibacterial agent. Plates were incubatedunder aerobic conditions at room temperature for 7 days. Plates supplementedwith chloramphenicol which yielded significant growth of bacteria, as determinedby colony morphology, were excluded from further analysis. Plates were examinedby eye for bacterial colonies or fungal foci whose morphological characteristicswere similar to previously characterized food-derived microbes. Candidatefood-derived microbes were isolated and identified by Sanger sequencing of the16S rRNA gene (for bacteria; primers used were 27f, 5- AGAGTTTGATCCTGGCTCAG, and1492r, 5-GGTTACCTTGTTACGACTT) or ITS region (for fungi; primers used were ITS1f,5-CTTGGTCATTTAGAGGAAGTAA, and ITS4, 5-TCCTCCGCTTATTGATATGC). After selectcolonies had been picked for isolation, the surface of each plate was scrapedwith a razor blade to collect all remaining colonies, and material was suspendedin PBS. Dilutions were pooled, and DNA was extracted from the resulting pooledmaterial using a PowerSoil kit (MoBio, Carlsbad, CA). The remaining pooledmaterial was stocked in 20% glycerol and stored at−80°C.
Supplementary Material
Short-term diet alters the gut microbiota
Ten subjects were tracked across each diet arm. (A) Fiberintake on the plant-based diet rose from a median baseline value of9.3±2.1 to 25.6±1.1 g/1,000kcal (p=0.007; two-sided Wilcoxonsigned-rank test), but was negligible on the animal-based diet (p=0.005).(B) Daily fat intake doubled on the animal-based diet from abaseline of 32.5±2.2% to 69.5±0.4% kcal(p=0.005), but dropped on the plant-based diet to 22.1±1.7%(p=0.02). (C) Protein intake rose on the animal-based diet to30.1±0.5% kcal from a baseline level of16.2±1.3% (p=0.005) and decreased on the plant-based diet to10.0±0.3% (p=0.005). (D) Within-sample speciesdiversity (α-diversity, Shannon’s Diversity Index), did notsignificantly change during either diet. (E) The similarity of eachindividual’s gut microbiota to their baseline communities(β-diversity, Jensen-Shannon distance) decreased on the animal-baseddiet (dates with q<0.05 identified with asterisks; Bonferroni-corrected,two-sided Mann-Whitney U test). Community differences were apparent one dayafter a tracing dye showed the animal-based diet reached the gut (blue arrowsdepict appearance of food dyes added to first and last diet day meals; Extended Data Fig.3a).
Bacterial cluster responses to diet arms
Cluster log2 fold-changes on each diet arm were computed relative tobaseline samples across all subjects and are drawn as circles. Clusters withsignificant fold-changes on the animal-based diet are colored in red, andclusters with significant fold-changes on both the plant- and animal-based dietsare colored in both red and green. Uncolored clusters exhibited no significantfold-change on either the animal or plant-based diet (q<0.05, two-sidedWilcoxon signed-rank test). Bacterial membership in the clusters with the threelargest positive and negative fold-changes on the animal-based diet are alsodisplayed and colored by phylum: Firmicutes (purple), Bacteroidetes (blue),Proteobacteria (green), Tenericutes (red), and Verrucomicrobia (gray). MultipleOTUs with the same name are counted in parentheses.
Diet alters microbial activity and gene expression
Fecal concentrations of SCFAs from (A) carbohydrate and(B) amino acid fermentation (*p<0.05, two-sidedMann-Whitney U test; n=9–11 fecal samples/diet arm; Supplementary Table 11).The animal-based diet was associated with significant increases in geneexpression (normalized to reads per kilobase per million mapped, or RPKM;n=13–21 datasets/diet arm) among (C) glutamineamidotransferases (K08681, vitamin B6 metabolism), (D)methyltransferases (K00599, polycyclic aromatic hydrocarbon degradation), and(E) beta-lactamases (K01467). (F) Hierarchicalclustering of gut microbial gene expression profiles collected on theanimal-based (red) and plant-based (green) diets. Expression profile similaritywas significantly associated with diet (p<0.003; two-sidedFisher’s exact test excluding replicate samples), despiteinter-individual variation that preceded the diet (Extended Data Figs.6a,b). Enrichment on animal-based diet (red) and plant-based diet (green)for expression of genes involved in (G) amino acid metabolism and(H) central metabolism. Numbers indicate the mean fold-changebetween the two diets for each KEGG orthologous group assigned to a givenenzymatic reaction (Supplementary Table 17). Enrichment patterns on the animal- andplant-based diets agree perfectly with patterns observed in carnivorous andherbivorous mammals, respectively2 (p<0.001,Binomial test). Note: Pyr Cx is represented by two groups, which showeddivergent fold-changes. Asterisks in panels C-E andG,H indicate p<0.05, Student’st test. Values in panels A-E are mean±sem.Abbreviations: glutamate dehydrogenase (GDH), glutamate decarboxylase (Glu Dx),succinate-semialdehyde dehydrogenase (SSADH), phosphoenolpyruvate carboxylase(PEPCx), pyruvate carboxylase (Pyr Cx), phosphotransferase system (PTS), PEPcarboxykinase (PEPCk), oxaloacetate decarboxylase (ODx), pyruvate,orthophosphate dikinase (PPDk).
Foodborne microbes are detectable in the distal gut
(A) Common bacteria and fungi associated with theanimal-based diet menu items, as measured by 16S rRNA and ITS gene sequencing,respectively. Taxa are identified on the genus (g) and species (s) level. A fulllist of foodborne fungi and bacteria on the animal-based diet can be found inSupplementary Table21. Foods on the plant-based diet were dominated by matches to theStreptophyta, which derive from chloroplasts within plant matter (Extended Data Fig. 7a).(B-E). Fecal RNA transcripts were significantly enriched(q<0.1, Kruskal-Wallis test; n=6–10 samples/diet arm) forseveral food-associated microbes on the animal-based diet relative to baseline(BL) periods, including (B) Lactococcus lactis,(C) Staphylococcus carnosus, (D)Pediococcus acidilactici, and (E) aPenicillium sp. A complete table of taxa with significantexpression differences can be found in Supplementary Table 22. (F) Fungalconcentrations in feces before and 1–2 days after the animal-based dietwere also measured using culture media selective for fungal growth (plate countagar with milk, salt, and chloramphenicol). Post-diet fecal samples exhibitsignificantly higher fungal concentrations than baseline samples(p<0.02; two-sided Mann-Whitney U test; n=7–10 samples/dietarm). (G) Increased RNA transcripts from the plant-derived Rubus chlorotic mottle virus transcripts increase on the plant-based diet (q<0.1, Kruskal-Wallistest; n=6–10 samples/diet arm). Barplots (B-G) all displaymean±sem.
Changes in the fecal concentration of bile acids and biomarkers for Bilophilaon the animal-based diet
(A) Deoxycholic acid, a secondary bile acid known topromote DNA damage and hepatic carcinomas26, accumulates significantly on the animal-based diet(p<0.01, two-sided Wilcoxon signed-rank test; see Supplementary Table 23for the diet response of other secondary bile acids). (B) RNA-Seqdata also supports increased microbial metabolism of bile acids on theanimal-based diet, as we observe significantly increased expression of microbialbile salt hydrolases (K01442) during that diet arm (q<0.05,Kruskal-Wallis test; normalized to reads per kilobase per million mapped, orRPKM; n=8–21 samples/diet arm). (C) Total fecal bile acidconcentrations also increase significantly on the animal-based diet, relative tothe preceding baseline period (p<0.05, two-sided Wilcoxon signed-ranktest), but do not change on the plant-based diet (Extended Data Fig. 9).Bile acids have been shown to cause IBD in mice by stimulating the growth of thebacterium Bilophila6, which is known to reduce sulfite to hydrogen sulfide viathe sulfite reductase enzyme (dsrA; Extended Data Fig. 10). (D) Quantitative PCRshowed a significant increase in microbial DNA coding for dsrA on theanimal-based diet (p<0.05; two-sided Wilcoxon signed-rank test), and(E) RNA-Seq identified a significant increase in sulfitereductase expression (q<0.05, Kruskal-Wallis test; n=8–21samples/diet arm). Barplots (B,E) display mean±sem.
Footnotes
Supplementary Information
Supplementary information is linked to the online version of the paperat www.nature.com/nature.
Author contributions
LAD, RJD, and PJT designed the study, and developed and prepared thediets. LAD, CFM, RNC, DBG, JEB, BEW, and PJT performed the experimental work.AVL, ASD, YV, MAF, and SBB conducted bile acid analyses. LAD and PJT performedcomputational analyses. LAD and PJT prepared the manuscript.
RNA-Seq data are deposited in the Gene Expression Omnibus (GEO) database(accession number GSE46761); 16S and ITS rRNA gene sequencing reads aredeposited in MG-RAST (project ID 6248). Reprints and permissions information areavailable at www.nature.com/reprints. The authors have no competinginterests.
Acknowledgements
We would like to thank Andrew Murray, Guido Guidotti, Erin O’Shea,Jeffrey Moffitt, and Bodo Stern for insightful comments; Mary Delaney (HarvardDigestive Disease Core) for biochemical analyses; Christian Daly, Michele Clamp, andClaire Reardon for sequencing support; N. Fierer for providing ITS primers; An Luong and Kylynda Bauer for technicalassistance; Jennifer Brulc and Ravi Menon (General Mills) for nutritionalguidelines; Atiqur Rahman for menu suggestions; Aviva Must and Jeanette Queenan fornutritional analysis; and our diet study volunteers for their participation. Thiswork was supported by the National Institutes of Health (P50 GM068763), the BostonNutrition Obesity Research Center (DK0046200), and the General Mills Bell Instituteof Health and Nutrition, Minneapolis, MN.