Review Article

Role of Gut Microbiota in the Aetiology of Obesity: Proposed Mechanisms and Review of the Literature

Table 3

Evidence from animal studies about the role of gut microbiota in obesity.  

ReferenceStudy modelAim of the studyStudy design and outcomes measuresResultsConclusion

Studies suggesting association of gut microbiota with obesity

Fleissner et al. 2010 [14]Male adult C3H GF and CV miceInfluence of different diets on the body composition of GF and CV miceAd libitum intake of low fat (LF), high fat (HF), and commercial Western diet (WD) for GF and CV mice. Real-time PCR, FISH, and fiaf/angplt4 in gut and bloodGF mice gained more weight and body fat and had less energy expenditure than CV mice on HF. Higher Firmicutes (especially Erysipellotrichacae) and lower Bacteroides in CV mice on HF and WD. Intestinal Fiaf increased in GF mice but no change in plasma fiaf levels as compared to CV mice GF mice are not protected from diet induced obesity. Diet affects gut microbiota composition and fiaf does not play a role in fat storage mediated by gut microbiota

Šefčíková et al. 2010 [31]8–10 pups per nest, Sprague-Dawley rats, from day 21 to day 40Effect of normal and overnutrition on the development of gut microbiota, intestinal alkaline phosphatase, and occurrence of obesityStandard laboratory diet for control group and additional milk based liquid diet for study group. Bacterial enumeration via FISH, alkaline phosphatase activity via immunocytochemistryObese rats gained more energy (25%) and higher body fat (27%) than lean rats. Alkaline phosphatase increased in obese rats. Lactobacilli increased while Bacteroides decreased in obese rats significantly This study may provide a baseline for further insight into the ways of involvement in programming of a sustained intake and digestion

Ding et al. 2010 [32]GF/CONV mice and NF-κB knockin mice (GF/CONV)Hypothesis: intestinal inflammation is promoted by the interaction of gut bacteria and high fat diet, contributing to the progression of insulin resistance and obesityHigh and low fat diets for 2, 6, or 16 weeks. GF mice fed with diet after exposure to faecal slurries of CONV mice. Blood glucose and ELISA for insulin. TNF-α mRNA expression by qPCR. Expression of NFkB mice by fluorescent light microscopyCONV mice gained more weight than GF. Increased expression of TNF-α mRNA and NF-κB in CONV HF diet mice. TNF-α changes precede weight changes. Enhanced NF-κB in GF NF-κB mice on feeding CONV NFkB faecal slurryHF diet and enteric bacteria interact to promote inflammation and insulin resistance prior to the development of weight gain, adiposity, and insulin resistance

Turnbaugh et al. 2008 [17]8-9-week-old GF/CONV miceTo study the interrelationship between diet, energy balance, and gut microbiota using mouse model of obesityConventionalisation of GF mice with HF Western diet followed by introduction of Western or CHO diet in CONV mice. CARB-reduced or FAT-reduced diets in another subset. qPCR, DEXA scan, and weight measurements doneWestern diet-associated caecal community had a significantly higher relative abundance of the Firmicutes (specifically Mollicutes) and lower Bacteroidetes. Mice on the Western diet gained more weight than mice maintained on the CHO diet and had significantly more epididymal fat. Mice on CARB-R and FAT-R diet consumed fewer calories, gained less weight, and had less fatThere is restructuring of gut microbiota with Western diet, specifically reduction of Bacteroides and surge in Mollicutes class of Firmicutes with increased capacity to harvest energy from diet

Daniel et al. 2014 [33]Male C57BL/6NCrl mice (, per group)To investigate changes in function and activity of the gut ecosystem in response to dietary changeLC-MS/MS for metaproteome, FT-ICR-MS for metabolome, Miseq illumina pyrosequencing. Intervention with high fat (HF) and control (carbohydrate) diet for 12 weeks HF diet did not affect caecal taxa richness. Bacterial communities clustered according to diet. Significantly ↓ Ruminococcaceae (Firmicutes) and ↑ Rikenellaceae (phylum Bacteroidetes), Lactobacilli, and Erysipelotrichiales in HF fed versus carbohydrate fed diet. 19 OTUs affected by HF diet. Carbohydrate and HF group had distinct proteome and metabolomeHigh fat diet affects gut microbial ecology both in terms of composition and function

Cani et al. 2007 [9]C57bl6/J mice 
and 
CD14−/−
mutant strain
To evaluate the influence of gut microbiota on the development of metabolic endotoxemiaMetabolic, inflammatory, and 
microbiological differences 
(by FISH) between high fat fed 
obese or rodent lean chow-fed 
mice
High fat feeding and obesity decimate intestinal microbiota– Bacteroides-mouse intestinal bacteria, Bifidobacterium, and Eubacterium rectale-Clostridium coccoides groups all significantly ↓ compared to in control animalsHigh fat diet induces changes in gut microbiota that leads to elevated plasma LPS leading to metabolic endotoxemia, by altering the gut barrier function

Cani et al. 2008 [34]C57bl6/J ob/ob
mice
Manipulating gut microorganisms through antibiotics to demonstrate whether changes in gut microbiota control the occurrence of metabolic syndromes Caecal microbiota of mice under 
high fat low fibre diet and antibiotics. qPCR and DGGE
Antibiotic reduced LPS caecal content and metabolic endotoxemia in both ob/ob and high fat groups. High fat diet ↑ intestinal permeability and LPS uptake leading to metabolic endotoxemia. Absence of CD14 mimicked the metabolic and inflammatory effects of antibioticsHigh fat diet modifies gut microbiota which induce inflammation and metabolic endotoxemia. Antibiotics can reverse these changes

Murphy et al. 2010 [16]HF fed wild-type mice and leptin deficient ob/ob mice ( per group)To investigate the effect of high fat diet and genetically determined obesity for changes in gut microbiota and energy harvesting capability over timeGC, metagenomic pyrosequencing 
high fat or normal chow diet fed to ob/ob mice and wild-type mice for 8 weeks
↑ in Firmicutes and Bacteroidetes in HF fed and obese mice but not in lean. Changes in microbiota not associated with markers of energy harvest. Initial increase in caecal SCFA (acetate) and ↓ in stool energy with HF diet did not remain significant over time Changes in bacterial phyla are a function of high fat diet and are not related to the markers of energy harvest

de Wit et al. 2012 [35]Male C57BL/6J miceTo study the effect of dietary fat type (polyunsaturated and saturated fatty acids ratio) on the development of obesityPhylogenetic microarray (MITChip) analysis, bomb calorimetry, measurement of triglycerides, and plasma insulinHF diet with high saturated fatty acids (palm oil) induced ↑ weight gain and liver TG compared to HF diet with olive oil and safflower oil. HF diet with palm oil ↓ microbial diversity and ↑ Firmicutes (Bacilli, Clostridium clusters XI, XVII, and XVIII) Bacteroidetes ratio. Upregulation of 69 lipid metabolism genes in distal small intestine and ↑ fat in stool Type of dietary fat influences the weight gain and hepatic lipid metabolism

Faith et al. 2011 [36]Male C57BL/6J mice ( per group)Changes in 10 model gut communities species’ abundance and microbial genes with changes in peculiar diet Shotgun sequencing of faecal DNA 
diets used for each community: casein (for protein), corn oil (for fat), starch (for polysaccharides), and sucrose (for simple sugars)
61% variance in abundance of the community members was explained by diet particularly casein. Absolute abundance of E. rectale, Desulfovibrio piger, and M. formatexigens ↓ by 25–50% while Bacteroides caccae ↑ with increase in casein, although the total community biomass ↑Host diet explains configuration of gut microbiota both for refined diets and complex polysaccharides

Hildebrandt et al. 2009 [37]RELM-β knockout female mice and wild-type miceTo assess the influence of host phenotype, genotype, immune function, and diet on gut microbiota 16S rDNA 454 FLX pyrosequencing, metagenomic sequencingSwitching to high fat diet caused ↓ Bacteroidetes and ↑ Firmicutes and Proteobacteriain both wild-type and RELM-β knockout mice irrespective of the genotype. Genetic makeup only modestly influenced the gut microbiome composition. Changes in gene content with HF dietDiet determines the gut microbiota composition

Huang et al. 2013 [38]Adult male C57BL/6To assess the relationship of diet content and source on gut microbiota and adiposity16S rRNA analysis, terminal restriction fragment length polymorphism and V3-V4 sequence tag analysis via next generation sequencing. Mesenteric fat and gonadal fat tissue analysis.
Milk, lard, or safflower based diets for 4 weeks
↑ weight gain and caloric intake with HF compared to low fat diet. Milk based and PUFA based diets animals had ↑ adipose tissue inflammation than lard based or low fat diet. Milk based and PUFA diet had significantly ↑ Proteobacteria and ↓ Tenericutes. PUFA based fed animals had ↑ expression of adipose tissue inflammation genes (MCP1, CD192, and resistin)Dietary fat components reshape gut microbiota and alter adiposity and inflammatory status of the host

Jakobsdottir et al. 2013 [39]Male Wister ratsTo investigate the effect of dietary fibre on metabolic risk markers in low and high fat diets at 2, 4, and 6 weeksGas liquid chromatography, liver fat content, cholesterol and triglycerides analysis, and terminal fragment length polymorphism. Diets supplemented with guar gum or a mixture↓ in weight gain, liver fat, cholesterol, and triglycerides with fibre. Change in formation of SCFA. ↓ in serum SCFA with HF diet followed by recovery after 4 weeks. Succinic acid ↑ with HF consumption. Dietary fibre ↓ this effect and also ↓ inflammation. Bacteroides were ↑ with guar gum and Akkermansia was ↑ with fibre-free dietHF diet ↑ metabolic risk factors which are partly reversed by high fibre diet

Studies suggesting effect of diet on changes in gut microbiota and resultant obesity

de la Serre et al. 2010 [15]Male Sprague-Dawley ratsTo evaluate whether changes in gut bacteria and gut epithelial function are diet or obese associatedIntestinal permeability, intestinal Alk-Pase, plasma LPS, tissue myeloperoxidase (MPO) activity, immunochemical localization of TLR4/MD2 complex, and Occludin. Sequence analysis of the microbial 16S rRNA gene Appearance of two distinct groups; diet induced obesity prone (DIO-P) and diet induced obesity resistant (DIO-R) groups. DIO-P rats had ↑ features of adiposity, ↑ MPO activity, ↑ TLR4 MD2 immunoreactivity and ↑ plasma LPS levels, ↑ gut permeability, immunoreactivity of Occludin, and ↓ alkaline phosphatase levels than LF and DIO-R group. HF diet was associated with ↑ Clostridiales regardless of propensity for obesity. A marked difference in Enterobacteriales in DIO-P animals compared with either DIO-R or LF fed animalsChanges in gut bacteria are independent of obese status. Gut inflammation marked by increased LPS may be a triggering mechanism for hyperphagia and obesity

Bäckhed et al. 2004 [1] Adult germ-free (GF) C57BL/6 miceTo evaluate the effect of gut microbiota on the host energy metabolism using animal modelConventionalisation of GF mice with murine gut microbiota or B. thetaiotaomicron, intestinal fiaf, liver metabolism, total body fat, LPL activity in adipose tissue, and faecal microbiota composition by qPCR Conventionalized GF mice showed 57% ↑ in body fat, increased energy expenditure, ↓ intestinal fiaf, increased LPL activity, and ↑ expression of ChREBP and SREBP-1 in liver. Firmicutes to Bacteroides ratio similar in GF and CONVGut microbiota alter host energy storage by affecting fiaf and LPL activity

Bäckhed et al. 2007 [13]Adult GF C57BL/6 mice () and CONV mice ()To assess whether GF mice are protected against obesity on high fat Western dietDietary intervention with low fat followed by high fat Western diet for 8 weeksCONV mice gained ↑ weight on HF diet while conventionalised GF mice did not. Stool energy was similar to the LF fed GF mice. Persistent ↑ TG in HF fed GF mice. GF mice had ↑ Acc-p, AMPK-P, and Cpt-1 activity. GF mice had ↓ hepatic glycogen and glycogen-synthase activity. ↑ fiaf in HF fed GF mice GF mice are protected against diet induced obesity by two mechanisms: () increased phosphorylated AMPK and () increased fiaf

Vijay-Kumar et al. 2010 [40]TLR5 knockout mice (T5KO), wild-type mice (WT)To show that mice deficient in TLR-5 exhibit hyperphagia, which is a principal factor in the development of obesity and metabolic syndromeBroad spectrum antibiotics. Pyrosequencing of 16S rRNA genes in the caecum. Transplantation of TLR5-KO mice microbiota into WT germ-free hostsAntibiotic treatment ↓ the bacterial load by 90%, correction of metabolic syndrome similar to the wild-type mice. Relative abundance of bacterial phyla was similar in both, with 54% Firmicutes, 39.8% Bacteroides. 116 phyla observed to be enriched or ↓ in TLR5-KO relative to WT mice. Microbiota of WT mice transplanted to the TLR5-KO mice resulted in all features of metabolic syndrome in the TLR5-KO groupLoss of TLR-5 results in metabolic syndrome and alteration in gut microbiota

Ley et al. 2005 [3]Leptin deficient C57BL/6J ob/ob mice, lean ob/+, and +/+ mice ()To study differences in bacterial diversity between obese genetic model of obesity and its relationship with kinship16S rRNA gene amplification of caecal bacteria followed by analysis using PHRED and PHRAP software. All mice fed the same polysaccharide rich chowob/ob mice consumed 42% more chow and gained significantly ↑ weight. Mothers and offspring shared bacterial community. Obese ob/ob mice had 50% reduction in Bacteroidetes and a proportional ↑ in Firmicutes as compared to lean regardless of the kinship and genderObesity is associated with altered bacterial ecology. This however needs to be correlated with the metabolic attributes of gut microbial diversity

Turnbaugh et al. 2006 [4]Leptin deficient C57BL/6J ob/ob mice () and lean ob/+ and +/+ mice ()Whether gut microbial gene content correlates with characteristic distal gut microbiome of leptin deficient ob/ob mice and their lean counterparts 1S rRNA whole genome shotgun metagenomics, GC-MS for SCFA analysis, bomb calorimetry, gut microbiota transplantation, and DEXAFirmicutes-enriched obese microbiome clustered together while lean phenotype with ↓ Firmicutes to Bacteroidetes ratio clustered together. Obese microbiome rich in enzymes for breakdown of dietary polysaccharides particularly glycoside hydrolases. ob/ob had ↑ acetate and butyrate and significantly ↓ stool energyObese microbiome is associated with increased energy harvest

Caesar et al. 2010 [12]Swiss-Webster mice (GF, CONV, and E. coli monocolonised mice)Whether gut microbiota especially LPS promote inflammation in white adipose tissue (WAT) and impair glucose metabolism DEXA, insulin, and glucose tolerance. Macrophage isolation, immunohistochemistry, and flow cytometry and immunoblot in WAT, LPS analysis, and RT-qPCRMonocolonisation of GF mice with E. coli W3110 or isogenic strain MLK1067 with low immunogenic LPS had impaired glucose tolerance. However, only GF mice with E. coli W3110, and not MLK1067, showed ↑ proinflammatory macrophage infiltration in WATMacrophage accumulation is microbiota dependent but impaired glucose tolerance is not

Caricilli et al. 2011 [41]TLR2 knockout mice (TLR2−/−) and wild-type mice ( per group)Influence of gut microbiota on metabolic parameters, glucose intolerance, insulin sensitivity, and insulin signalling in TLR2 knockout mice454 pyrosequencing↑ Firmicutes (47.92% versus 13.95%), Bacteroidetes (47.92% versus 42.63%), and ↓ Proteobacteria (1.04% versus 39.53%) in TLR2−/−. ↑ LPS absorption, insulin resistance, impaired insulin signalling, and glucose intolerance in TLR2−/− compared to controlsAlteration in gut microbiota in non-germ-free conditions links genotype to phenotype

Everard et al. 2013 [42]C57BL/6 mice (genetically obese, HF fed, and type-2 diabetic)To ascertain the role of Akkermansia muciniphila in obesity and type-2 diabetesReal-time qPCR, MITChip analysis, LTO-Orbitrap mass spectrometer, and ELISA for insulin and faecal IgAAkkermansia muciniphila ↓ obesity and type-2 diabetes which was normalised by oligofructose. Administration of A. muciniphila reversed markers of metabolic disorders. These effects needed viable A. muciniphilaThis microorganism could be used as part of a potential strategy for the treatment of obesity

Fei and Zhao 2013 [43]C57BL/6J GF miceEndotoxin producing Enterobacter cloacae B29 isolated from obese human gut could induce obesity and insulin resistance in GF mice16S rRNA gene sequencing for bacteria and limulus amebocyte lysate test for endotoxin measurementMonocolonisation of GF mice with E. cloacae induced obesity and insulin resistance on HF diet while GF control mice only on HF diet did not. Enterobacter-colonised GF obese mice had ↑ plasma endotoxin levels and inflammatory markers Gut microbiota-produced endotoxin may be causatively related to obesity in human hosts

Geurts et al.
2011 [20]
Leptin resistant db/db miceTo investigate the gut microbiota composition in obese and diabetic leptin resistant mice versus lean miceCombined pyrosequencing and phylogenetic microarray analysis of 16S rRNA gene↑ Firmicutes, Proteobacteria, and Fibrobacteres phyla in db/db mice compared to lean mice. Odoribacter, Prevotella, and Rikenella were exclusively present in db/db mice while Enterorhabdus was identified exclusively in lean mice. db/db mice had ↑ tone of eCB and ↑ apelin and APJ mRNA levels Gut microbiota vary with genotype and play a significant role in the regulation of eCB and apelin/APJ mRNA system

GF: germ-free mice, CV: conventionally raised germ-free mice, HF: high fat diet, LF: low fat diet, WD: Western diet, PCR: polymerase chain reaction, FISH: florescent in situ hybridization, fiaf/angptl4: fasting induced adipocyte factor/angiopoietin-like-protein factor-4, NF-κB: nuclear factor-kappaB, CHO: carbohydrate, CARB-R: carbohydrate-reduced diet, FAT-R: FAT-reduced, DEXA or DXA: dual energy X-ray absorptiometry, FT-ICR-MS: Fourier-transform ion cyclotron resonance mass spectrometry, OTUs: operational taxonomic units, LPS: lipopolysaccharide, DGGE: denaturing gradient gel electrophoresis, GC: gas chromatography, SCFA: short chain fatty acids, RELM-β: resistin-like molecule-β, PUFA: polyunsaturated fatty acids, MCP1: monocyte chemoattractant protein 1, Alk-Pase: alkaline phosphatase, TLR4/MD2: Toll-Like Receptor 4/mitogen detector-2, ChREBP: carbohydrate response element binding protein, SERBP-1: sterol response element binding protein-1, TG: triglycerides, Cpt-1: carnitine palmitoyltransferase-1, AMPK: adenosine monophosphate kinase-1, Acc-p: acetyl CoA carboxylase (phosphorylated), WT: wild-type, GC-MS: gas chromatography-mass spectrometry, and eCB: endocannabinoid receptor system.