The Human Gut Microbiome and Body Metabolism: Implications ...pages.ucsd.edu/~mboyle/COGS163/pdf-files/The Human... · cation of whole-genome sequencing technologies for metagenomic

The Human Gut Microbiome and Body Metabolism:Implications for Obesity and Diabetes

Sridevi Devaraj,1,2 Peera Hemarajata,1,2 and James Versalovic1,2*

BACKGROUND: Obesity, metabolic syndrome, and type 2diabetes are major public health challenges. Recently,interest has surged regarding the possible role of theintestinal microbiota as potential novel contributors tothe increased prevalence of these 3 disorders.

CONTENT: Recent advances in microbial DNA sequenc-ing technologies have resulted in the widespread appli-cation of whole-genome sequencing technologies formetagenomic DNA analysis of complex ecosystemssuch as the human gut. Current evidence suggests thatthe gut microbiota affect nutrient acquisition, energyharvest, and a myriad of host metabolic pathways.

CONCLUSION: Advances in the Human MicrobiomeProject and human metagenomics research will leadthe way toward a greater understanding of the impor-tance and role of the gut microbiome in metabolic dis-orders such as obesity, metabolic syndrome, anddiabetes.© 2013 American Association for Clinical Chemistry

Obesity, metabolic syndrome, and type 2 diabetes aremajor public health challenges, affecting approxi-mately 26 million children and adults in the US. Morethan 8% of the US population has diabetes, of which17.9 million people have the metabolic syndrome (1 ).During the past 20 years, obesity has dramatically in-creased in prevalence in the US. More than 1 in 3 USadults (36%) are obese, and approximately 12.5 mil-lion (17%) of children and adolescents (age 2–19 years)are obese (2 ). In the US in 2010 (2 ), all of the states hada prevalence of obesity of over 20%. The heterogeneityof these disorders has been demonstrated through bothanthropometric and genetic studies. These metabolicdisorders are believed to be caused by a combination ofgenetic susceptibilities and lifestyle changes. Recently,interest has surged in the possible role of the intestinal

microbiome as a potential contributor to the rapidlyincreased prevalence of obesity (3–5 ). This review fo-cuses on recent advances in the understanding of thegut microbiome and techniques to assess the micro-biome and its relationship to human body metabolism,obesity, metabolic syndrome, and type 2 diabetes (Fig. 1).

The Human Gut Microbiome: The Toolkit behindthe Science

The widespread application of 16S rRNA gene se-quencing for detection of bacterial pathogens and mi-crobial ecology has provided a robust technical plat-form for the evaluation of the bacterial composition ofthe human microbiome. Sequencing of 2 primary tar-gets within bacterial 16S rRNA genes yielded valuablecompositional data pertaining to the human fecal mi-crobiome of 242 healthy adults (6, 7 ). In the HumanMicrobiome Project, 18 different body sites were sam-pled and sequenced. Stool specimens were the singlespecimen type used to study the intestinal microbiome.Previously published studies demonstrated the varia-tion in composition of the gut microbiome among lo-cations within the gastrointestinal tract in differentmammalian species. For example, 16S rRNA gene se-quencing has been deployed to study the maturationof murine cecal microbiota, and these studies dem-onstrated the existence of a large number of yet-unidentified bacteria that inhabit the mammalian in-testine (6 ). Such sequencing strategies, which are cul-ture independent, are essential for determiningbacterial composition of the microbiome and its rela-tive stability and diversity over time. Thus, it is essentialto develop robust experimental models of the humanmicrobiome to delineate important mechanistic pro-cesses in the development of human disease states.

Advances in sequencing technologies have resultedin the widespread application of whole-genome (WG)3

sequencing technologies for metagenomic DNA anal-

1 Department of Pathology and Immunology, Baylor College of Medicine and2 Department of Pathology, Texas Children’s Hospital, Houston, TX.

* Address correspondence to this author at: Department of Pathology, TexasChildren’s Hospital, 1102 Bates Ave., Suite 830, Houston, TX 77030. Fax832-825-1165; e-mail [email protected].

Received September 4, 2012; accepted January 16, 2013.Previously published online at DOI: 10.1373/clinchem.2012.187617

3 Nonstandard abbreviations: WG, whole genome; NMR, nuclear magnetic res-onance; GC-TOFMS, gas chromatography TOF mass spectrometry; SCFA, short-chain fatty acid; CAZymes, carbohydrate-active enzymes; CLA, conjugatedlinoleic acid; GLP-1, glucagon-like peptide 1; RYGB, Roux-en-Y gastric bypass;HILIC-HPLC, hydrophilic interaction liquid chromatography–HPLC; TNF, tumornecrosis factor; GABA, �-amino butyric acid; Fiaf, fasting-induced adipocytefactor; LPS, lipopolysaccharide; TLR, Toll-like receptor; BB, biobreeding.

Clinical Chemistry 59:4617–628 (2013) Review

617

ysis of complex ecosystems such as the human intestine(7 ). WG sequencing strategies provide microbial com-positional as well as functional information. WG datacan be used to infer bacterial composition, and thesedata yield information similar to that generated by 16SrRNA gene sequencing. The genome sequences ofhighly abundant species are well represented in a set ofrandom shotgun reads, whereas less abundant speciesare represented by fewer sequences generated in a next-generation sequencing run. This relative richness per-mits the comprehensive measurement of the composi-

tional responses of an ecosystem to dietary changes,drug therapy, epigenetic alterations, and environmen-tal perturbations. Alternatively, most genes (usuallyapproximately 2000 genes per bacterium) in the micro-biome are sequenced so that metabolic and other func-tional pathways can be evaluated in each individual’smetagenome. Functional WG data provide opportuni-ties to find out which metabolic pathways are affectedand how the microbiome may contribute mechanisti-cally to health and disease states. This technology cre-ates the formidable challenge of managing vast data

Fig. 1. Hyperglycemia (HG) and increased free fatty acids (FFA), which are hallmarks of obesity, metabolicsyndrome, and diabetes, combined with a high-fat, high–glycemic load diet, could result in increased activation ofthe inflammasome complex as well as increase the activation of macrophages via increased TLR activation andnuclear factor �B (NF-�B) activation.

Increased metabolic endotoxemia may occur and activate the TLR4 pathway via the adapter protein, MyD88, leading to immunecell activation and inflammation. Also, macrophages could infiltrate the adipose tissue and activate mitogen-activated proteinkinases, such as c-Jun aminoterminal kinase (JNK) and NF-�B, resulting in increased cross-talk and adipose-tissue–derivedadipokines. A hyperglycemic and high fat diet could also result in changes to the gut microbiome by altering the content ofhistidine, glutamate, SCFAs, and other factors and promote gut-barrier dysfunction and conditions prevalent in obesity,metabolic syndrome, and diabetes by altering the host response. All of these metabolic alterations that result in increasedsystemic inflammation, macrophage activity, and TLR activation contribute to the increased cardiometabolic burden in obesity,diabetes, and metabolic syndrome.

Review

618 Clinical Chemistry 59:4 (2013)

sets. Advances in next-generation DNA sequencingyielded 576.7 Gb of microbial DNA sequence data,which were generated with an Illumina™ genome an-alyzer (Illumina) from total DNA from the stool sam-ples of 124 European adults (8 ). The relationship be-tween the commensal microbiota that comprise the gutmicrobiota and those that are in the intestinal barrier iscomplex and differs spatially throughout different ar-eas of the gastrointestinal tract. Fecal metagenomicsmeasures ecosystem changes in stool or the distal intes-tine, but it does not compare the microbiomes in dif-ferent regions of the intestine. It is also important tonote that metagenomic analysis of fecal samples doesnot include all important molecular interactionswithin the gastrointestinal tract. Turnbaugh et al. haveproposed the idea of a core set of functions within themicrobiome, and the tools of proteomics and metabo-lomics may be required for more in-depth functionalanalyses (7, 9 ). From a systems perspective, meta-genomic analyses may provide further details on spe-cific intraindividual changes and thus have major im-plications for personalized medicine strategies.

Metatranscriptomics, metaproteomics, and meta-bonomics will be useful to explore the functional as-pects of the gut microbiome from the top down. Real-time analysis of the intestinal microbiome is a usefultool in the development of personalized approaches totargeted therapies. Metabonomics can be described asthe study of metabolic responses to chemicals, the en-vironment, and diseases and involves the computa-tional analysis of spectral metabolic data that provideinformation on temporal changes to specific metabo-lites. In addition, metabonomics provides global met-abolic profiling of an individual in real time. It is pos-sible, with such approaches, to elucidate complexpathways and networks that are altered in specific dis-ease states. The combination of metabolic profiling andmetagenomic studies of gut microbiota permits thestudy of host and microbial metabolism in great detail.Such analysis of functional components of the micro-biome that affect metabolism and human health is re-ferred to as functional metagenomics.

Metagenomics and the science of the human mi-crobiome have arrived at the forefront of biology pri-marily because of major technical and conceptual de-velopments. The major technical development was thedeployment in many centers of next-generation DNAsequencing technologies with greatly enhanced capa-bilities for sequencing collections of microbial ge-nomes in the metagenome. Technological advanceshave created new opportunities for the pursuit of large-scale sequencing projects that were difficult to imaginea decade ago. The key conceptual development was theemerging paradigm of the essential nature of complexmicrobial communities and their importance to mam-

malian biology and human health and disease. The Hu-man Microbiome Project was approved in May 2007 as1 of 2 major components (in addition to the humanepigenomics program) of NIH RoadMap version 1.5(now known as the Common Fund). Recently, 2 sem-inal reports from the Human Microbiome Project con-sortium (10, 11 ) described investigations in which apopulation of 242 healthy adults were sampled at 15 or18 body sites up to 3 times, 5177 microbial taxonomicprofiles were generated from 16S rRNA genes, andmore than 3.5 T bases of metagenomic sequences weregenerated. In addition, in parallel, the Human Micro-biome Project consortium has sequenced approxi-mately 800 human-associated reference genomes. Thisresource will provide a framework for future studies ofdisease states and a reference collection of healthy humanmicrobiome data. The data set will enable future investi-gations into the epidemiology and ecology of the humanmicrobiome in various disease states, and treatment strat-egies will evolve from these studies. Using compositionaland functional approaches, the relationships betweenpathological variations in the gut microbiome and severaldisease states have been delineated.

Urine metabolomics provides an opportunity forstudies of the microbiome’s impact on whole-bodymetabolism. The advantages of using urinary samplesinclude relatively large sample volumes and the conve-nience of noninvasive collection. In addition, urinesamples can be used for the investigation of the chro-nology of metabolic changes and thus are a valuabletool for investigations related to the pathogenesis orprogression of disease and for screening and diagnosisas well as prognostic evaluation. The methods com-monly used for metabolic profiling of urine includeprocedures such as nuclear magnetic resonance(NMR) spectroscopy, LC-MS, GC-MS, and gas chro-matography TOF mass spectrometry (GC-TOFMS). Ina recent seminal report, the Nicholson group describeda method for urine collection and storage that empha-sizes the importance of midstream urine collection andthe addition of urease before the freezing of urine sam-ples. This method will eventually be used for metabolicprofiling. Before analyses by GC-MS– based tech-niques, urease activity is terminated with ethanol ormethanol and then derivatized by subjecting the sam-ple to oximation followed by trimethylsilyl derivatiza-tion (12 ). Because of the various sample preparationsteps, it is important to use biological QC samples andcheck the validity of the data that are obtained fromGC-MS– based techniques that use principal compo-nent analysis. GC-MS– based metabolomic studies in-clude several steps such as baseline correction, noisereduction, deconvolution, peak area calculation, andretention time alignment, and these steps help to gen-erate consistent data. Several different commercially

The Human Gut Microbiome and Body Metabolism Review

Clinical Chemistry 59:4 (2013) 619

available software programs can assist in such correc-tion strategies before data analyses. Also, in urinarymetabolic profiling, it is important to normalize data(on the basis of volume, creatinine content, and othervariables) to obtain meaningful information and min-imize the influence of dilution. Using this protocol,investigators were able to undertake high-throughputmetabolic profiling of approximately 400 – 600 metab-olites in 120 urine samples per week. High-throughputanalyses by NMR spectroscopy or MS are metabolic-profiling strategies that are widely used to provideglobal metabolic overviews of human metabolism (13–16 ). Coupled with computational multivariate analy-sis, these methods provide a deeper understanding ofdisease states and can lead to biomarker discovery. Thisapproach facilitates the quantification of environmen-tal influences on the host genome and human health.As part of large-scale clinical studies, this analyticalstrategy has been successfully applied to disease statessuch as hypertension (17 ), ischemic heart disease (18 ),diabetes (19 ), and obesity (20 ).

Metabonomics can be quite challenging becausethe chemical space associated with the endogenous me-tabolites is highly diversified, and thus complete meta-bolic information for any sample is hard to deciphercompletely. Common analytical technologies used inmetabolomics and metabonomics include NMR spec-troscopy, LC-MS, and GC-MS, as well as GC-TOFMS.These different analytical techniques have their ownstrengths and weaknesses and are usually used in anintegrated fashion such that each of these analyticalplatforms can provide complementary data; the selec-tion of particular analytical techniques depends on thestudy questions that are being posed. NMR has the ad-vantage of being rapid, nondestructive to samples, andapplicable to intact biomaterials rich in chemical struc-tural information. NMR requires minimal samplepreparation and can be used to investigate a mixture ofor several different metabolites in a single sample.However, MS-based strategies have the advantages ofincreased sensitivity, accuracy, precision, and repro-ducibility compared to NMR. Furthermore, the cou-pling of GC to TOFMS offers several additional advan-tages such as reduced analysis time and greateraccuracy with respect to peak deconvolution.

The Gut Microbiome: Beyond Composition toFunction and Metabolism

The gut microbial community includes approximately1014 bacteria that normally reside in the gastrointesti-nal tract, reaching a microbial cell number that greatlyexceeds the number of human cells of the body. Thecollective genome of these microorganisms (the micro-biome) contains millions of genes (a rapidly expanding

number) compared to roughly 20 000 –25 000 genes inthe human genome. This microbial “factory” contrib-utes to a broad range of biochemical and metabolicfunctions that the human body could not otherwiseperform (21 ). Although diet-induced changes in gutmicrobiota occur within a short time frame (1–3– 4days after a diet switch), the changes are readily revers-ible (22, 23 ). In animal models, the ratio of the mostprominent intestinal bacterial phyla, the Bacteroidetesand Firmicutes, is altered in response to dietarychanges (22, 23 ). Disruption of the energy equilibriumleads to weight gain. Mouse model studies have dem-onstrated the relationship between energy equilibrium,diet, and the composition of the gut microbiome.Transplantation of the gut microbiota from obese do-nors resulted in increased adiposity in recipients com-pared to a similar transfer from lean donors.

Recent evidence suggests that the gut microbiotaaffect nutrient acquisition, energy harvest, and a myr-iad of host metabolic pathways (24 ). Recent findingsraise the possibility that the gut microbiota has an im-portant role in regulating weight and may be partlyresponsible for the development of obesity. Initial evi-dence of the relationship between obesity and gut mi-crobial composition was reported 3 decades ago, whensurgically induced weight loss through gastric bypasssurgery and weight gain through lesions of the ventro-medial hypothalamic nucleus were found to be associ-ated with changes in gut microbial ecology (25, 26 ).These earlier studies used culture-dependent methods,which detect a minority of microbes harbored in thegut. In recent years, the ability to obtain a thoroughpicture of gut microbial communities has improved bythe introduction of molecular, culture-independenttechniques based on ribosomal 16S rRNA gene se-quencing. Jumpertz et al. (27 ) performed an inpatientenergy balance study in 12 lean and 9 obese individualsas they consumed 2 calorically distinct diets for briefperiods of time, and these investigators simultaneouslymonitored the gut microbiota by performing pyrose-quencing studies of bacterial 16S rRNA genes presentin feces and by measuring ingested and stool calories bybomb calorimetry. This study showed that altered nu-trient load (i.e., high calories vs low calories) inducedrapid changes in the bacterial composition of the hu-man gut microbiota, and these changes correlated wellwith stool energy loss in lean individuals. Increasedproportions of Firmicutes and corresponding reduc-tions in Bacteroidetes taxa were associated with an in-creased energy harvest of approximately 150 kcal.These data point to a strong link between gut micro-biome composition and nutrient absorption in hu-mans, and such studies need to be confirmed withlarger numbers of study participants.

Review


The gut microbiome is very important in main-taining both gastrointestinal and immune function aswell as being crucial for the digestion of nutrients, andthis notion has been confirmed by studies of germ-freemice (28 –30 ). Important metabolic functions of thegut microbiome include the catabolism of dietary tox-ins and carcinogens, synthesis of micronutrients, fer-mentation of indigestible food substances, and assist-ing in the absorption of electrolytes and minerals. Inaddition, the production of short-chain fatty acids(SCFAs) by the gut microbiome affects growth and dif-ferentiation of enterocytes and colonocytes. Differ-ences in the metabolic activities of the gut microbiomemay contribute to variation in caloric extraction fromingested dietary substances, storage of calories in adi-pose tissue, and energy availability for microbial pro-liferation. Such differences in the gut microbiome arealso responsible for the variation in the ability of anindividual’s capacity to harvest energy, which may ex-plain aspects of obesity. Differences in gut microbialcomposition and its metabolic efficiency may be re-sponsible for the predisposition of an individual tometabolic disorders such as obesity and diabetes (31 ).

The gut microbiome can affect whole-body me-tabolism and alter physiological parameters in multiplebody compartments (32 ). In one study (33 ), gnotobi-otic mice had increased quantities of phosphocholineand glycine in the liver and increased quantities of bileacids in the intestine. The gut microbiome also influ-ences kidney homeostasis by modulating quantities ofkey cell-volume regulators such as betaine and choline(33 ). A more recent study showed specific differencesin the patterns of bile acids present and reduced overallbile acid diversity in germ-free vs conventional rats(34 ). Compared to conventional rats, germ-free ratshave increased concentrations of conjugated bile acidsthat can accumulate in the liver and the heart.

The Gut Microbiome and CarbohydrateMetabolism

Carbohydrates are an important nutritional compo-nent for mammals and the mammalian microbiome,including the gut microbiota. Mammals absorb simplesugars, including galactose and glucose, in the proxi-mal jejunum via specific sugar transporters. Mamma-lian enzymes hydrolyze disaccharides (sucrose, lactose,maltose) and starches to constituent monosaccharides,but have limited abilities to hydrolyze other polysac-charides. As a consequence, every day a bulk of undi-gested plant polysaccharides (cellulose, xylan, and pec-tin) and partially digested starch reaches microbialcommunities in the distal gut. By hosting metabolicallyactive microbiota capable of hydrolyzing complex car-bohydrates, mammals avoid the need to evolve com-

plex enzymes that are required to break down the vari-ety of polysaccharides in the diet. Microbes, bycontrast, contain many genes encoding a variety ofcarbohydrate-active enzymes (CAZymes) in the hu-man microbiome (35 ). Microbial CAzymes that con-stitute the mammalian host repertoire include glyco-side hydrolases, carbohydrate esterases, glycosyltransferases, and polysaccharide lyases (35 ). Microbesgain access to abundant readily fermentable carbonsources that would otherwise be wasted by the host andmay use these complex carbohydrate substrates to sus-tain viable, functionally robust microbial communitiesand generate bioactive signals that affect mammalianmetabolism.

Intestinal bacterial taxa differ with respect to theirabilities to utilize dietary and host-derived carbohy-drates (e.g., mucus components) (23, 36 ). Bacterio-detes (23, 36 ) also have been demonstrated to easilyassimilate dietary carbohydrates, because members ofthis bacterial phylum possess several carbohydrate uti-lization pathways. However, in situations of dietarycarbohydrate starvation, gut bacteria catabolize mu-cins in the gastrointestinal tract as a carbohydratesource, thereby potentially compromising the mucuslayer adjacent to the epithelium. In addition to Bacte-roides, strains of the genus Bifidobacterium containgenes encoding glycan-foraging enzymes that enablethese gut bacteria to acquire nutrients from host-derived glycans (37 ). Besides their capacity to hydro-lyze starch, gut microbes have developed the ability todegrade numerous plant and host-derived glycoconju-gates (glycans) and glycosaminoglycans including cel-lulose, chondroitin sulfate, hyaluronic acid, mucins,and heparin. Microbial catabolic enzymes such as en-doglycosidases may act on dietary substrates to releasecomplex N-glycans from human milk and other dairysources (38 ). Fluctuations in diet may have functionalconsequences for bacteria and the host so that the “can-nibalization” of indigenous mammalian carbohydratesmay result in augmentation of beneficial features, pre-vention of diseases, or predisposition to different dis-ease states. For example, bifidobacteria grown on hu-man milk oligosaccharides stabilize tight junctionformation in the epithelium and promote the secretionof the antiinflammatory cytokine, interleukin-10 (39 ).The biogeography of the microbiome may be relevantbecause specific genes/pathways such as simple carbo-hydrate transport phosphotransferase systems aremore prominent in the small intestine than in the colon(40 ). Probing into pathways which are affected by al-terations in the gut microbiome (e.g., carbohydratestorage and utilization) will yield new knowledge onthe role of human-associated microbes in the develop-ment of several metabolic disorders in humans.



The Gut Microbiome and Fatty Acid Metabolism

Intestinal bacteria including probiotics produce a di-verse array of fatty acids that may have health-promoting effects. Intestinal bifidobacteria produceconjugated linoleic acid (CLA), and CLA appears tomodulate the fatty acid composition in the liver andadipose tissue in murine models (41 ). In addition toconjugated and free fatty acids, intestinal bacteria gen-erate SCFAs (i.e., acetate, butyrate, propionate) by fer-menting dietary carbohydrates (fiber) that humanscannot digest themselves. A recent study showed thatgerm-free mice are devoid of SCFAs, indicating the im-portance of the gut microbiota for SCFA production inthe intestine (42 ). Acetate is the dominant SCFA typein humans, and this SCFA appears to play an intriguingrole in the modulation of 5�AMP-activated protein ki-nase activity and macrophage infiltration in adiposetissue (43 ). SCFAs such as propionate can be used forde novo glucose or lipid synthesis and serve as an en-ergy source for the host.

SCFAs may function as microbe-derived signalsthat influence carbohydrate metabolism and gut phys-iology by stimulating mammalian peptide secretionand serving as energy sources for gut epithelial cells.SCFAs can stimulate glucagon-like peptide 1 (GLP-1)secretion via the G-protein– coupled receptor FFAR2(free fatty acid receptor 2) in the colonic mucosa (44 ).By stimulating GLP-1 secretion, bacterial SCFAs pro-vide signals that suppress glucagon secretion, induceglucose-dependent insulin secretion, and promote glu-cose homeostasis. An enteroendocrinological pathwayis proposed in which SCFAs stimulate the secretion ofpeptide YY, a hormone that is released by ileal andcolonic epithelial cells in response to feeding and seemsto suppress the appetite (45 ). High-fat diets supple-mented with butyrate prevented and reversed insulinresistance in dietary-obese mice. At the same time,butyrate-producing bacteria and fecal butyrate con-centrations decline with diets containing reducedamounts of specific carbohydrates (46 ). The SCFApropionate modulates energy homeostasis by promot-ing GPR41 (G protein-coupled receptor 41)-mediatedactivation of sympathetic neurons, in contrast to ke-tone bodies (47 ). The ability to modulate sympatheticoutflow provides another mechanism linking the gutmicrobiome to the enteric nervous system, energy ex-penditure, and metabolic homeostasis.

Roux-en-Y gastric bypass (RYGB) surgery is a ma-jor bariatric intervention to treat morbid obesity. Be-fore surgery, increased quantities of Bacteroidetes wereobserved, but reductions in Bacteroidetes and en-hanced quantities of Proteobacteria were detected fol-lowing surgery (48 ). These microbial population shiftslikely change the metabolite profiles and relative pre-

ponderance of different fatty acids, including SCFAs.These results are supported by a recent animal study.Nonobese rats with RYGB had decreased amounts ofFirmicutes and Bacteroidetes and significantly in-creased amounts (52-fold higher concentrations) ofProteobacteria compared with sham-operated rats(49 ). Obesity is a proinflammatory state. It wasshown that the abundance of butyrate-producingFaecalibacterium prausnitzii species is negatively as-sociated with biomarkers of inflammation beforeand after RYGB, indicating that this bacterial speciesmay contribute to maintaining a healthy gut (48 ).Thus, surgical interventions in the gastrointestinaltract may have profound effects on gut microbialcomposition, SCFA production, and the mamma-lian immune system.

Interestingly, a recent study (50 ) has demon-strated that subtherapeutic administration of antibiot-ics alters the population structure of the gut micro-biome as well as its metabolic capabilities. In this study,investigators administered subtherapeutic doses of an-tibiotics to young mice, resulting in increased adiposityin young mice and increased levels of the incretinGIP-1. In addition, these investigators observed sub-stantial taxonomic changes in the microbiome (in-creased Lachnospiraceae and Firmicutes and decreasedBacteroidetes), changes in key genes involved in themetabolism of carbohydrates to SCFAs (increased lev-els of acetate, propionate, and butyrate), increases incolonic SCFA levels, and alterations in the regulation ofhepatic metabolism of lipids and cholesterol. Thus,modulation of murine metabolic homeostasis can beachieved by altering the gut microbiota through anti-biotic manipulation.

The Gut Microbiome and Amino Acid Metabolism

Beneficial microbes such as bifidobacteria and lactoba-cilli produce biologically active compounds derivedfrom amino acids, including a variety of biogenicamines. Dietary components include proteins and pep-tides that may be hydrolyzed to amino acids by luminalproteinases and peptidases. Amino acids derived fromdietary protein sources may serve as substrates for lu-minal bioconversion by the gut microbiome. Diversemicrobial enzymes may contribute to mammalianamino acid metabolism by generating bioactive metab-olites in the intestine. One such class of enzymes,amino acid decarboxylases, is widely prevalent in gutmicrobes, and these microbial enzymes, when com-bined with amino acid transport systems, link dietarycompounds with microbial metabolism and signalingwith the gut mucosa (Fig. 2).

Combinations of metabolomics strategies, includ-ing MS, HPLC, and NMR, are leading to discoveries of

Review


metabolites and small compounds derived from thehuman microbiome. With the use of hydrophilic inter-action liquid chromatography–HPLC (HILIC-HPLC),bioactive molecules derived from gut microbes wereisolated as antiinflammatory, tumor necrosis factor(TNF)-inhibitory compounds and HILIC-HPLC frac-tions were analyzed by MS and NMR. One such mi-crobial signal and biogenic amine, histamine, wasidentified and quantified in TNF-inhibitory HILIC-HPLC fractions derived from Lactobacillus reuterifound in breast milk and the gut (51 ). Histamine isproduced from L-histidine via histidine decarboxyl-ase, which is present in some fermentative bacteriaincluding probiotic lactobacilli. One constituent ofthe gut microbiome, L. reuteri, is able to convert adietary component, L-histidine, into an immuno-regulatory signal, histamine, which suppresses pro-inflammatory TNF production via histamine type 2receptors in the intestinal epithelium. Other exam-ples of microbe-facilitated amino acid metabolisminclude the generation of �-amino butyric acid(GABA) from glutamate via glutamate decarboxyl-ase (52 ) and the production of putrescine fromornithine. The identification of these bacterial bio-active metabolites and their corresponding mecha-nisms of action with respect to immunomodulationmay lead to improved antiinflammatory strategiesfor chronic immune-mediated diseases. Such antiin-flammatory amino acid metabolites may amelioratepathologic processes in obesity and diabetes.

The Gut Microbiome and Body Metabolism:Obesity and Inflammation

The incidence of overweight and obesity has reachedepidemic proportions. Data reported by the CDC andthe National Health and Nutrition Examination Sur-vey indicated that, in 2008, an estimated 1.5 billionadults were overweight, and more than 200 millionmen and almost 300 million women were obese bythese criteria. Worldwide obesity has more than dou-bled in the last 2 decades. Obesity is associated with acluster of metabolic and systemic disorders such as in-sulin resistance, type 2 diabetes, fatty liver disease, ath-erosclerosis, and hypertension. The major cause ofobesity is a positive energetic balance resulting from anincreased energy intake from the diet and a decreasedenergy output associated with low physical activity. Inaddition to alterations in diet and physical activity re-sulting in obesity, genetic differences contribute toobesity and cause differences in energy storage and ex-penditure. Furthermore, growing evidence suggeststhat the gut microbiota represents an important factorcontributing to the host response to nutrients. A land-mark study by Turnbaugh et al. (53 ) was one of the firststudies to show how the gene content in the gut micro-biota contributes to obesity. The microbiomes ob-tained from the distal gut of genetically obese leptin-deficient mice (ob/ob) and their lean littermates (ob/�and �/�) were compared. In this study, investigatorsreported that the microbiota in the ob/ob mice con-

Fig. 2. Intestinal microbes may play an important role in host-microbiota interactions via luminal conversion.

Nutrients consumed by the host may be converted by intestinal microbes into several bioactive compounds that could affectthe health and disease states of the host and the intestinal microbiota. SCFAs � short-chain fatty acids. Reproduced withpermission from Hemarajata et al. (78 ).



tained genes encoding enzymes that hydrolyze indi-gestible dietary polysaccharides. Increased amounts offermentation end products (such as acetate and bu-tyrate) and decreased calories were found in the feces ofobese mice. These data suggest that the gut microbiotain this mouse model promoted the extraction of addi-tional calories from the diet.

The composition of the gut microbiome seems tobe important in regulating body weight (54 ). To dem-onstrate this point, investigators conducted experi-ments in which they transplanted the gut microbiota ofeither ob/ob mice or lean mice to lean gnotobioticmice. After 2 weeks, mice that received microbiotafrom the ob/ob mice were able to extract more caloriesfrom food and also showed a significantly greater fatgain than mice that received the microbiota from leanmice. Thus, differences in caloric extraction of ingestedfood substances may be largely a result of the compo-sition of the gut microbiota. These data support a piv-otal role for the gut microbiome in the pathogenesis ofobesity and obesity-related disorders.

Manipulation of the gut microbiota may be an im-portant therapeutic strategy to regulate energy balancein individuals who are obese, diabetic, or have a diag-nosis of metabolic syndrome. In genetically obese,ob/ob mice and their lean counterparts fed the samepolysaccharide-rich diet, Ley et al. analyzed bacterial16S rRNA gene sequences from the cecal microbiotaand reported that ob/ob mice had 50% fewer Bacte-roidetes and correspondingly more Firmicutes thantheir lean littermates and this difference was unrelatedto differences in food consumption.

Backhed et al. (55 ) confirmed these findings andfound that young, conventionally reared mice had a40% higher body fat content and 47% higher gonadalfat content than germ-free mice, although their foodconsumption was less than their germ-free counter-parts. When the distal gut microbiota from young,conventionally-reared mice were transplanted into thegnotobiotic mice, a 60% increase in body fat within 2weeks was observed, without any increase in food con-sumption or energy expenditure. This increase in bodyfat was accompanied by increased insulin resistance,adipocyte hypertrophy, and increased concentrationsof circulating leptin and glucose. Mechanistic studiesdemonstrated that the microbiota promoted absorp-tion of monosaccharides from the gut and induced he-patic lipogenesis in the host. These responses werelargely mediated via upregulation of 2 signaling pro-teins, ChREBP (carbohydrate response element-binding protein) and liver SREBP-1 (sterol responseelement-binding protein type-1). In addition, with ge-netically modified fasting-induced adipocyte factor(Fiaf)-knockout mice, gut microbes were also shown tosuppress intestinal Fiaf (56 ).

Several studies have highlighted the pivotal role ofinflammation in the metabolic processes leading to themetabolic syndrome, obesity, and diabetes. Cani et al.(57–59 ) postulated another mechanism linking the in-testinal microbiota to the development of obesity. Theauthors hypothesized that bacterial lipopolysaccharide(LPS) derived from gram-negative bacteria residing inthe gut microbiota may be the trigger for increased in-flammation observed in high-fat diet–induced metabolicsyndrome. In a series of experiments in mice fed a high-fatdiet, the investigators showed evidence of pronouncedendotoxemia, associated with reductions in both gram-negative (Bacteroides-related bacteria) and gram-positivebacteria (Eubacterium rectale—Clostridium coccoidesgroup and bifidobacteria), and an increased ratio ofgram-negative to gram-positive bacteria. The authorsof this report suggested that chronic metabolic endo-toxemia induces obesity, insulin resistance, anddiabetes.

In human experiments, Ley et al. (60 ) and Ravus-sin et al. (61 ) serially monitored the fecal gut microbi-ota in 12 obese individuals who participated in aweight-loss program for a year by following either afat-restricted or carbohydrate-restricted low-caloriediet. Similar to experiments in mice, in humans a rela-tive abundance of microbiota that belonged to the Bac-teroidetes and Firmicutes phyla was found, and the mi-crobiota showed remarkable intraindividual stabilityover time. Before the initiation of the low-calorie diet, arelative abundance of Firmicutes and decreasedamounts of Bacteroidetes were detected in the obeseparticipants compared with the nonobese controls. Af-ter weight loss, increased amounts of Bacteroidetes(3% to 15%) and a decreased abundance of Firmicuteswere observed, and these changes correlated with thepercentage of weight loss and not with changes in di-etary caloric content. These human studies confirmedanimal data suggesting that alterations in gut microbialcomposition are associated with obesity. Cause and ef-fect relationships between obesity and changes in thegut microbiota remain unclear. Kalliomäki et al., in aprospective study of children from birth to age 7 years(62 ), collected fecal specimens at 6 and 12 months ofage. This report documented an abundance of Bifido-bacterium taxa and decreased proportions of Staphylo-coccus aureus in children who were whose weight waswithin reference intervals at age 7 years than in thosewho were overweight or obese. Although they did notexamine factors such as diet and physical activity, thesedata suggest that differences in the composition of thegut microbiota precede overweight and obesity status.Antibiotics have also been shown to pervasively affectgut microbial composition. A 5-day course of orallyadministered ciprofloxacin decreased substantially thediversity of the fecal microbial community (63 ). In this

Review


study, although most of the microbial community re-vived within 4 weeks after administration of cipro-floxacin, some other genera failed to reappear even af-ter treatment with the antibiotic for 6 months (63 ).

The Gut Microbiome and Metabolism: Diabetes andthe Metabolic Syndrome

Toll like receptors (TLRs) are pattern recognition re-ceptors that are important in mediating inflammationand immunity. Increased amounts of TLRs are presenton cell surfaces in patients with obesity, diabetes, andmetabolic syndrome (63 ). Recently, investigators haveexplored the role of the gut microbiome in regulatingTLR-mediated insulin resistance. Mice deficient in themicrobial pattern-recognition receptor TLR5 dis-played hyperphagia, became obese, and developed fea-tures of the metabolic syndrome, including hyperten-sion, hypercholesterolemia, and insulin resistancesecondary to dysregulation of interleukin-1� signaling(43 ). When gut microbiomes from these mice weretransplanted into germ-free mice with an intact toll-like receptor 5 (TLR5) gene, recipient mice developedsimilar features of the metabolic syndrome, which sug-gests that the intestinal microbiome was the key deter-minant of this disease phenotype. In another study,TLR2-deficient mice developed obesity, insulin resis-tance, and glucose intolerance, and the gut micro-biomes of TLR2-deficient mice had a greater abun-dance of Firmicutes and fewer Actinobacteria of thegenus Bifidobacterium (64 ). Administration of an anti-biotic cocktail eliminated many of the Firmicutes andresulted in improved insulin activity and glucose toler-ance. In addition to improved insulin activity and glu-cose tolerance, because lower levels of Bifidobacteriumspp. contribute to increased gut permeability, thischange in the gut microbiome may result in a leaky gutand yield increased concentrations of endotoxins suchas LPS in the circulation. The immune system recog-nizes LPS as a microbial pattern triggering TLR signal-ing and causing inflammation. Both obesity and in-flammation tend to cause diabetes, so the loss of TLR2in these mice leads to changes in their gut bacteria,which result in a greater risk of diabetes mellitus. In-deed, increased amounts of TLR2 have been observedon monocytes of patients with metabolic syndrome,type 1 and type 2 diabetes compared with matchedcontrols (65–72 ). TLR2 deficiency in diabetic mice re-sults in decreased development of complications of di-abetes such as diabetic nephropathy (68 ). Possibly, thegut microbiome plays a crucial role in regulating dia-betic vasculopathies, and this area will be an importantarea of future investigation.

With regard to the role of the microbiome in met-abolic syndrome and associated abnormalities, studies

in germ-free mice demonstrate that they are protectedfrom obesity, insulin resistance, dyslipidemia, and fattyliver disease/nonalcoholic steatohepatitis when fed ahigh-fat Western diet (69 ). In contrast, following thecolonization with microbiota from conventionallyraised mice, the body fat content in the originally germ-free mice increased up to 60% in 14 days. This wasassociated with increased insulin resistance, althoughthe food intake was reduced. The metabolic syndromeaffects 1 in 3 US adults and leads to an increased pro-pensity of diabetes and cardiovascular disease (70 ). In asingle human study in patients with the metabolic syn-drome, Zupancic et al. (71 ) studied Amish men andwomen with varying body mass indices. In 310 studyparticipants, gut microbiota were characterized bydeep pyrosequencing of bar-coded PCR ampliconsfrom the V1–V3 region of the 16S rRNA gene. Theywere able to identify 3 communities of interacting bac-teria in the gut microbiota, analogous to previouslyidentified gut enterotypes. Network analysis identified22 bacterial species and 4 operational taxonomic unitsthat were either positively or inversely correlated withmetabolic syndrome traits, suggesting that certainmembers of the gut microbiota can contribute to themetabolic syndrome. It is important that future studiesfocus on delineation of specific components of the gutmicrobiome that contribute to visceral obesity, dysgly-cemia, dyslipidemia, hypertension, and insulin resis-tance associated with this population (71 ). Nonalco-holic fatty liver disease is the hepatic manifestation ofmetabolic syndrome and the leading cause of chronicliver disease in the Western world. Using differentmouse models of inflammasome deficiency, such asmice deficient in Asc (apoptosis-associated speck-likeprotein containing a caspase recruitment domain),NLRP3 (nucleotide-binding domain, leucine rich fam-ily, pyrin containing 3), caspase, or interleukin 18, theauthors showed significant alterations in gut microbi-ota as evidenced by increased members of Bacterio-detes and decreased members of Firmicutes in thesemouse models. More severe hepatic steatosis and in-flammation were found, as evidenced by increasedTLRs (mainly TLR 4 and 9) and secretion of hepaticTNF-�. Importantly, the authors speculated that in-creased hepatic steatosis was due to intestinal bacterialproducts acting as agonists for TLR4 and 9 and enter-ing the liver via the portal circulation (72 ). Further-more, these pathologic changes result in exacerbationof hepatic steatosis and obesity (72 ). Thus, altered in-teractions between the gut microbiota and the host,produced by defective inflammasome sensing, maygovern the rate of progression of multiple abnormali-ties associated with metabolic syndrome.

Intestinal microbiomes have also been studied inrelation to insulin resistance in patients with type 2



diabetes. Larsen et al. reported that there was a signifi-cant reduction in the relative abundance of Firmicutesand Clostridia in adults with type 2 diabetes when theyused the technique of deep tag-encoded sequencing (73).Additionally, the ratios of Bacteroidetes to Firmicutes andBacteroides–Prevotella to C. coccoides–Eubacterium rectalegroups were correlated with increased fasting glucoselevels in these patients. In this study, Larsen et al.showed that in addition to the decreased abundance ofFirmicutes, the Betaproteobacteria levels were signifi-cantly increased in diabetic patients compared to non-diabetic controls and their abundance significantlycorrelated with their plasma glucose concentrations(r � 0.46, P � 0.05) (73 ). Such findings are intriguingand prompt questions regarding how microbial com-position and corresponding metabolites may influencewhole-body metabolism in humans and contribute toinsulin resistance and diabetes.

Interestingly, the gut microbiome also regulatestype 1 diabetes. Type 1 diabetes is an autoimmune dis-ease, which is due to the specific destruction of theendocrine insulin-secreting pancreatic � cells resultingin an insulinopenic state. Type 1 diabetes also predis-poses to microvascular and macrovascular complica-tions. Data have emerged on the critical role of thegastrointestinal microbiota in the protection or thetriggering of type 1 diabetes (74 ). In 2 models of dia-betes, the NOD (nonobese diabetic) mouse and thebiobreeding (BB) rat, the incidence of spontaneoustype 1 diabetes mellitus can be affected by the microbialenvironment in the animal-housing facility or by expo-sure to microbial stimuli (75 ). Furthermore, the recog-nition of bacterial determinants from intestinal micro-biota may trigger type 1 diabetes. TLRs are innatepattern-recognition receptors involved in host defenseand maintenance of tissue integrity. TLR signaling ismediated through the adapter protein MyD88, and thedeletion of MyD88 protects from atherosclerosis. Micelacking MyD88 were protected against insulitis (74 ),and this phenomenon depends on commensal mi-crobes because germ-free MyD88 knockout mice de-velop robust diabetes.

In BB rats (a model of type 1 diabetes), Lactobacil-lus species present in feces (L. johnsonii and L. reuteri)were negatively correlated with type 1 diabetes devel-opment (76 ), possibly via modulation of the intestinalmucosal protein and oxidative stress response leadingto lower quantities of proinflammatory cytokines suchas interferon �. Thus, therapeutic modulation aimed ataltering the gut microbiome may be beneficial in re-tarding the development of diabetes.

Alterations in the gut microbiota contribute to thedevelopment of autoimmune disorders such as type 1diabetes. Fecal samples were obtained from 4 pairs ofmatched participants in a case control study (77 ).

More than 30 billion nucleotide bases of Illumina shot-gun metagenomic data were analyzed, and the findingsrevealed significantly increased proportions of path-ways and modules involved in carbohydrate metabo-lism and stress responses in cases compared to con-trols. Other differences included the relative quantitiesof genes involved in adhesion, motility, and sulfur me-tabolism, which were more abundant in cases, whereasgenes with roles in DNA and protein metabolism,amino acid synthesis, and aerobic respiration weremore abundant in controls. The 16S rRNA data werealso mined for indications of changes in microbialcomposition. At the phylum level, numbers of Actino-bacteria, Bacteroidetes, and Proteobacteria were signif-icantly increased in cases, whereas Firmicutes, Fuso-bacteria, Tenericutes, and Verrucomicrobia werehigher in controls (P � 0.001). At the genus level, theabundance of Bacteroides was much greater in cases,whereas Prevotella was much more abundant in con-trols. Furthermore, the total number of bacteria thatproduce lactic acid and butyrate was greater in controlsthan in cases. Thus, these data suggest that such lactate-and butyrate-producing bacteria may be beneficial andmaintain a healthy gut and that dysregulation of thesebacteria can lead to reduction in optimal mucin syn-thesis, as identified in individuals with autoimmunediseases, and contribute to development of type 1 dia-betes. These exciting findings need to be confirmed inlarger study populations.

In conclusion, randomized clinical studies shouldhelp define the features of the gut microbiome thatcontribute to obesity and diabetes epidemics in definedpopulations. In addition, mechanistic studies of thehuman microbiome will be instructive and have ther-apeutic implications. Furthermore, advances in tech-nology, such as 16S rRNA sequencing, WG metag-enomics, and metabolomics in metabolic diseases willenable scientists to mine large data sets for disease-contributing features. Large databases and bioinfor-matics resources such as those derived from the Hu-man Microbiome Project will lead the way toward agreater understanding of the importance and role ofthe gut microbiome in metabolic disorders such asobesity, metabolic syndrome, and diabetes, and thesestudies may provide therapeutic strategies to reducethe aggregate cardiometabolic burden in humanpopulations.

Author Contributions: All authors confirmed they have contributed tothe intellectual content of this paper and have met the following 3 re-quirements: (a) significant contributions to the conception and design,acquisition of data, or analysis and interpretation of data; (b) draftingor revising the article for intellectual content; and (c) final approval ofthe published article.

Review


Authors’ Disclosures or Potential Conflicts of Interest: Upon man-uscript submission, all authors completed the author disclosure form.Disclosures and/or potential conflicts of interest:

Employment or Leadership: None declared.Consultant or Advisory Role: None declared.Stock Ownership: None declared.

Honoraria: None declared.

Research Funding: NIH P30 DK56338-06A2; J. Versalovic, NIHUH3 DK083990, NIH R01 AT004326-01A1, and NIH R01DK065075-01.

Expert Testimony: None declared.Patents: None declared.

References

1. Cowie CC, Rust KF, Ford ES, Eberhardt MS, Byrd-Holt DD, Li C, et al. Full accounting of diabetesand pre-diabetes in the U.S. population in 1988-1994 and 2005-2006. Diabetes Care 2009;32:287–94.

2. D’Adamo E, Santoro N, Caprio S. Metabolic syn-drome in pediatrics: old concepts revised, newconcepts discussed. Pediatr Clin North Am 2011;58:1241–55, xi.

3. Backhed F. Host responses to the human micro-biome. Nutr Rev 2012;70(Suppl 1):S147.

4. Tehrani AB, Nezami BG, Gewirtz A, Srinivasan S.Obesity and its associated disease: a role formicrobiota? Neurogastroenterol Motil 2012;24:305–11.

5. Harris K, Kassis A, Major G, Chou CJ. Is the gutmicrobiota a new factor contributing to obesityand its metabolic disorders? J Obes 2012;2012:879151.

6. Kibe R, Sakamoto M, Hayashi H, Yokota H, BennoY. Maturation of the murine cecal microbiota asrevealed by terminal restriction fragment lengthpolymorphism and 16S rRNA gene clone libraries.FEMS Microbiol Lett 2004;235:139–46.

7. Turnbaugh PJ, Gordon JI. The core gut micro-biome, energy balance and obesity. J Physiol2009;587:4153–8.

8. Caporaso JG, Lauber CL, Walters WA, Berg-LyonsD, Lozupone CA, Turnbaugh PJ, et al. Globalpatterns of 16S rRNA diversity at a depth ofmillions of sequences per sample. Proc Natl AcadSci U S A 2010;108(Suppl 1):4516–22.

9. Turnbaugh PJ, Hamady M, Yatsunenko T, Can-tarel BL, Duncan A, Ley RE, et al. A core gutmicrobiome in obese and lean twins. Nature2009;457:480–4.

10. The Human Microbiome Project Consortium.Structure, function and diversity of the healthyhuman microbiome. Nature 2012;486:207–14.

11. The Human Microbiome Project Consortium. Aframework for human microbiome research. Na-ture 2012;486:215–21.

12. Chan EC, Pasikanti KK, Nicholson JK. Global uri-nary metabolic profiling procedures using gaschromatography-mass spectrometry. Nat Protoc2011;6:1483–99.

13. Verberkmoes NC, Russell AL, Shah M, Godzik A,Rosenquist M, Halfvarson J, et al. Shotgun meta-proteomics of the human distal gut microbiota.ISME J 2009;3:179–89.

14. Nicholson JK, Lindon JC, Holmes E.‘Metabonomics’: understanding the metabolic re-sponses of living systems to pathophysiologicalstimuli via multivariate statistical analysis of bi-ological NMR spectroscopic data. Xenobiotica1999;29:1181–9.

15. Wikoff WR, Anfora AT, Liu J, Schultz PG, LesleySA, Peters EC, Siuzdak G. Metabolomics analysisreveals large effects of gut microflora on mamma-lian blood metabolites. Proc Natl Acad Sci U S A

2009;106:3698–703.16. Nicholson JK, Connelly J, Lindon JC, Holmes E.

Metabonomics: a platform for studying drug tox-icity and gene function. Nat Rev Drug Discov2002;1:153–61.

17. Holmes E, Loo RL, Stamler J, Bictash M, Yap IK,Chan Q, et al. Human metabolic phenotype di-versity and its association with diet and bloodpressure. Nature 2008;453:396–400.

18. Brindle JT, Antti H, Holmes E, Tranter G, Nichol-son JK, Bethell HW, et al. Rapid and noninvasivediagnosis of the presence and severity of coro-nary heart disease using 1H-NMR-based metabo-nomics. Nat Med 2002;8:1439–44.

19. Dumas ME, Wilder SP, Bihoreau MT, Barton RH,Fearnside JF, Argoud K, et al. Direct quantitativetrait locus mapping of mammalian metabolicphenotypes in diabetic and normoglycemic ratmodels. Nat Genet 2007;39:666–72.

20. Stella C, Beckwith-Hall B, Cloarec O, Holmes E,Lindon JC, Powell J, et al. Susceptibility of humanmetabolic phenotypes to dietary modulation. JProteome Res 2006;5:2780–8.

21. Neish AS. Microbes in gastrointestinal health anddisease. Gastroenterology 2009;136:65–80.

22. Hooper LV, Gordon JI. Commensal host-bacterialrelationships in the gut. Science 2001;292:1115–8.

23. Sonnenburg ED, Zheng H, Joglekar P, Higginbot-tom SK, Firbank SJ, Bolam DN, Sonnenburg JL.Specificity of polysaccharide use in intestinal bac-teroides species determines diet-induced microbi-ota alterations. Cell 2010;141:1241–52.

24. Nicholson JK, Holmes E, Kinross J, Burcelin R,Gibson G, Jia W, Pettersson S. Host-gut microbi-ota metabolic interactions. Science 2012;336:1262–7.

25. Clement K. Bariatric surgery, adipose tissue andgut microbiota. Int J Obes 2011;35(Suppl 3):S7–15.

26. Calvani R, Miccheli A, Capuani G, Tomassini Mic-cheli A, Puccetti C, Delfini M, et al. Gutmicrobiome-derived metabolites characterize apeculiar obese urinary metabotype. Int J Obes2010;34:1095–8.

27. Jumpertz R, Le DS, Turnbaugh PJ, Trinidad C,Bogardus C, Gordon JI, Krakoff J. Energy-balancestudies reveal associations between gut mi-crobes, caloric load, and nutrient absorption inhumans. Am J Clin Nutr 2011;94:58–65.

28. Gravitz L. Microbiome: the critters within. Nature2012;485:S12–3.

29. Tilg H, Kaser A. Gut microbiome, obesity, andmetabolic dysfunction. J Clin Invest 2011;121:2126–32.

30. Faith JJ, McNulty NP, Rey FE, Gordon JI. Predict-ing a human gut microbiota’s response to diet ingnotobiotic mice. Science 2011;333:101–4.

31. Kallus SJ, Brandt LJ. The intestinal microbiota andobesity. J Clin Gastroenterol 2011;46:16–24.

32. Zhao L, Shen J. Whole-body systems approachesfor gut microbiota-targeted, preventive health-care. J Biotechnol 2010;149:183–90.

33. Claus SP, Tsang TM, Wang Y, Cloarec O, Skordi E,Martin FP, et al. Systemic multicompartmentaleffects of the gut microbiome on mouse meta-bolic phenotypes. Mol Syst Biol 2008;4:219.

34. Swann JR, Want EJ, Geier FM, Spagou K, WilsonID, Sidaway JE, et al. Systemic gut microbialmodulation of bile acid metabolism in host tissuecompartments. Proc Natl Acad Sci U S A 2010;108(Suppl 1):4523–30.

35. Cantarel BL, Lombard V, Henrissat B. Complexcarbohydrate utilization by the healthy humanmicrobiome. PLoS One 2012;7:e28742.

36. Sonnenburg JL, Xu J, Leip DD, Chen CH, WestoverBP, Weatherford J, et al. Glycan foraging in vivoby an intestine-adapted bacterial symbiont. Sci-ence 2005;307:1955–9.

37. Turroni F, Bottacini F, Foroni E, Mulder I, Kim JH,Zomer A, et al. Genome analysis of Bifidobacte-rium bifidum PRL2010 reveals metabolic path-ways for host-derived glycan foraging. Proc NatlAcad Sci U S A 2010;107:19514–9.

38. Garrido D, Nwosu C, Ruiz-Moyano S, Aldredge D,German JB, Lebrilla CB, Mills DA. Endo-beta-N-acetylglucosaminidases from infant-gut associ-ated bifidobacteria release complex N-glycansfrom human milk glycoproteins. Mol Cell Pro-teomics 2012;11:775–85.

39. Chichlowski M, De Lartigue G, German JB, Ray-bould HE, Mills DA. Bifidobacteria isolated frominfants and cultured on human milk oligosaccha-rides affect intestinal epithelial function. J PediatrGastroenterol Nutr 2012;55:321–7.

40. Zoetendal EG, Raes J, van den Bogert B, Aru-mugam M, Booijink CC, Troost FJ, et al. Thehuman small intestinal microbiota is driven byrapid uptake and conversion of simple carbohy-drates. ISME J 2012;6:1415–26.

41. O’Shea EF, Cotter PD, Stanton C, Ross RP, Hill C.Production of bioactive substances by intestinalbacteria as a basis for explaining probioticmechanisms: bacteriocins and conjugated linoleicacid. Int J Food Microbiol 2012;152:189–205.

42. Martin FP, Wang Y, Sprenger N, Yap IK, Lundst-edt T, Lek P, et al. Probiotic modulation of sym-biotic gut microbial-host metabolic interactions ina humanized microbiome mouse model. Mol SystBiol 2008;4:157.

43. Carvalho BM, Guadagnini D, Tsukumo DM,Schenka AA, Latuf-Filho P, Vassallo J, et al. Mod-ulation of gut microbiota by antibiotics improvesinsulin signalling in high-fat fed mice. Diabetolo-gia 2012;55:2823–34.

44. Tolhurst G, Heffron H, Lam YS, Parker HE, HabibAM, Diakogiannaki E, et al. Short-chain fattyacids stimulate glucagon-like peptide-1 secretionvia the G-protein-coupled receptor FFAR2. Diabe-tes 2012;61:364–71.



45. Zhou J, Hegsted M, McCutcheon KL, Keenan MJ,Xi X, Raggio AM, Martin RJ. Peptide YY andproglucagon mRNA expression patterns and reg-ulation in the gut. Obesity (Silver Spring) 2006;14:683–9.

46. Gao Z, Yin J, Zhang J, Ward RE, Martin RJ, LefevreM, et al. Butyrate improves insulin sensitivity andincreases energy expenditure in mice. Diabetes2009;58:1509–17.

47. Kimura I, Inoue D, Maeda T, Hara T, Ichimura A,Miyauchi S, et al. Short-chain fatty acids andketones directly regulate sympathetic nervoussystem via G protein-coupled receptor 41(GPR41). Proc Natl Acad Sci U S A 2011;108:8030–5.

48. Furet JP, Kong LC, Tap J, Poitou C, Basdevant A,Bouillot JL, et al. Differential adaptation of hu-man gut microbiota to bariatric surgery-inducedweight loss: links with metabolic and low-gradeinflammation markers. Diabetes 2010;59:3049–57.

49. Li JV, Ashrafian H, Bueter M, Kinross J, Sands C,le Roux CW, et al. Metabolic surgery profoundlyinfluences gut microbial-host metabolic cross-talk. Gut 2011;60:1214–23.

50. Cho I, Yamanishi S, Cox L, Methe BA, Zavadil J, LiK, et al. Antibiotics in early life alter the murinecolonic microbiome and adiposity. Nature 2012;30:621–6.

51. Thomas CM, Hong T, van Pijkeren JP, HemarajataP, Trinh DV, Hu W, et al. Histamine derived fromprobiotic lactobacillus reuteri suppresses TNF viamodulation of PKA and ERK signaling. PLoS One2012;7:e31951.

52. Kim JY, Chung EJ, Kim JH, Jung KY, Lee WY.Response to steroid treatment in anti-glutamicacid decarboxylase antibody-associated cerebel-lar ataxia, stiff person syndrome and polyendo-crinopathy. Mov Disord 2006;21:2263–4.

53. Turnbaugh PJ, Ley RE, Mahowald MA, Magrini V,Mardis ER, Gordon JI. An obesity-associated gutmicrobiome with increased capacity for energyharvest. Nature 2006;444:1027–31.

54. Ley RE, Backhed F, Turnbaugh P, Lozupone CA,Knight RD, Gordon JI. Obesity alters gut microbialecology. Proc Natl Acad Sci U S A 2005;102:11070–5.

55. Backhed F, Manchester JK, Semenkovich CF, Gor-don JI. Mechanisms underlying the resistance todiet-induced obesity in germ-free mice. Proc NatlAcad Sci U S A 2007;104:979–84.

56. Greiner T, Backhed F. Effects of the gut microbi-

ota on obesity and glucose homeostasis. TrendsEndocrinol Metab 2011;22:117–23.

57. Delzenne NM, Neyrinck AM, Backhed F, Cani PD.Targeting gut microbiota in obesity: effects ofprebiotics and probiotics. Nat Rev Endocrinol2011;7:639–46.

58. Cani PD, Delzenne NM. Involvement of the gutmicrobiota in the development of low grade in-flammation associated with obesity: focus on thisneglected partner. Acta Gastroenterol Belg 2010;73:267–9.

59. Cani PD, Delzenne NM. Interplay between obesityand associated metabolic disorders: new insightsinto the gut microbiota. Curr Opin Pharmacol2009;9:737–43.

60. Ley RE. Obesity and the human microbiome. CurrOpin Gastroenterol 2009;26:5–11.

61. Ravussin Y, Koren O, Spor A, LeDuc C, Gutman R,Stombaugh J, et al. Responses of gut microbiotato diet composition and weight loss in lean andobese mice. Obesity 2011;20:738–47.

62. Kalliomaki M, Collado MC, Salminen S, Isolauri E.Early differences in fecal microbiota compositionin children may predict overweight. Am J ClinNutr 2008;87:534–8.

63. Dethlefsen L, Relman DA. Incomplete recoveryand individualized responses of the human distalgut microbiota to repeated antibiotic perturbation.Proc Natl Acad Sci U S A 2010;108(Suppl 1):4554–61.

64. Caricilli AM, Picardi PK, de Abreu LL, Ueno M,Prada PO, Ropelle ER, et al. Gut microbiota is akey modulator of insulin resistance in TLR 2knockout mice. PLoS Biol 2011;9:e1001212.

65. Dasu MR, Devaraj S, Park S, Jialal I. Increasedtoll-like receptor (TLR) activation and TLR ligandsin recently diagnosed type 2 diabetic subjects.Diabetes Care 2010;33:861–8.

66. Jialal I, Huet BA, Kaur H, Chien A, Devaraj S.Increased toll-like receptor activity in patientswith metabolic syndrome. Diabetes Care 2012;35:900–4.

67. Creely SJ, McTernan PG, Kusminski CM, FisherfM, Da Silva NF, Khanolkar M, et al. Lipopolysac-charide activates an innate immune system re-sponse in human adipose tissue in obesity andtype 2 diabetes. Am J Physiol Endocrinol Metab2007;292:E740–7.

68. Devaraj S, Tobias P, Kasinath BS, Ramsamooj R,Afify A, Jialal I. Knockout of toll-like receptor-2attenuates both the proinflammatory state of di-abetes and incipient diabetic nephropathy. Arte-

rioscler Thromb Vasc Biol 2011;31:1796–804.69. Frazier TH, DiBaise JK, McClain CJ. Gut microbi-

ota, intestinal permeability, obesity-induced in-flammation, and liver injury. JPEN J Parenter En-teral Nutr 2011;35:14S20S.

70. Alberti KG, Eckel RH, Grundy SM, Zimmet PZ,Cleeman JI, Donato KA, et al. Harmonizing themetabolic syndrome: a joint interim statement ofthe International Diabetes Federation Task Forceon Epidemiology and Prevention; National Heart,Lung, and Blood Institute; American HeartAssociation; World Heart Federation; Interna-tional Atherosclerosis Society; and InternationalAssociation for the Study of Obesity. Circulation2009;120:1640–5.

71. Zupancic ML, Cantarel BL, Liu Z, Drabek EF, RyanKA, Cirimotich S, et al. Analysis of the gut micro-biota in the old order Amish and its relation tothe metabolic syndrome. PLoS One 2012;7:e43052.

72. Henao-Mejia J, Elinav E, Jin C, Hao L, Mehal WZ,Strowig T, et al. Inflammasome-mediated dysbio-sis regulates progression of NAFLD and obesity.Nature 2012;482:179–85.

73. Larsen N, Vogensen FK, van den Berg FW, NielsenDS, Andreasen AS, Pedersen BK, et al. Gut mi-crobiota in human adults with type 2 diabetesdiffers from non-diabetic adults. PLoS One 2010;5:e9085.

74. Wen L, Ley RE, Volchkov PY, Stranges PB, Avane-syan L, Stonebraker AC, et al. Innate immunityand intestinal microbiota in the development oftype 1 diabetes. Nature 2008;455:1109–13.

75. Pozzilli P, Signore A, Williams AJ, Beales PE. NODmouse colonies around the world–recent factsand figures. Immunol Today 1993;14:193–6.

76. Valladares R, Sankar D, Li N, Williams E, Lai KK,Abdelgeliel AS, et al. Lactobacillus johnsonii N6.2mitigates the development of type 1 diabetes inBB-DP rats. PLoS One 2010;5:e10507.

77. Brown CT, Davis-Richardson AG, Giongo A, GanoKA, Crabb DB, Mukherjee N, et al. Gut micro-biome metagenomics analysis suggests a func-tional model for the development of autoimmu-nity for type 1 diabetes. PLoS One 2011;6:e25792.

78. Hemarajata P, Versalovic J. Effects of probioticson gut microbiota: mechanisms of intestinal im-munomodulation and neuromodulation. TherapAdv Gastroenterol. 2013;6:39–51.

Review


Documents

The Human Gut Microbiome and Body Metabolism: Implications ...pages.ucsd.edu/~mboyle/COGS163/pdf-files/The Human... · cation of whole-genome sequencing technologies for metagenomic