F Bäckhed, Wallenberg Laboratory, University of Gothenburg, S-413 45 Gothenburg, Sweden. E-mail: Fredrik.Backhed@wlab.gu.se. Phone: +46-31-342-7833. Fax: +46-31-82-3762.
The human gut is home to vast numbers of bacteria (gut microbiota), which outnumber the cells in the human body by an order of magnitude. The gut microbiota has coevolved with humans and can be considered an organ of similar size as the liver, containing more than 1,000 cell types (bacterial species) and encoding 150-fold more genes than are present in the human genome. Accordingly, the gut microbiota may have profound effects on various host responses, either directly or indirectly, by modifying food components or endogenously produced molecules into signaling molecules. Recent findings suggest that an altered gut microbial composition is associated with inflammatory bowel disease and obesity, indicating that the gut microbiota should be considered a contributing factor in several common diseases.
The human and mouse intestines exhibit significant postnatal development. For example, at birth, the blood vasculature is not fully developed, and Paneth cells, which produce antimicrobial peptides, are not present. The newborn gut is immediately colonized with bacteria from the mother and the environment, and these microorganisms may affect the developmental process. The gut microbiota also affects host physiology and metabolism in adulthood and may thus constitute an environmental factor. The development of germ-free mice, which are born and raised under axenic conditions, provides a powerful tool to investigate the underlying molecular mechanisms by which the gut microbiota affects host physiology. This review will focus on recent data demonstrating that the absence of a gut microbiota has profound effects on both intestinal and extraintestinal physiology.
THE GERM-FREE INTESTINE
The small intestines of germ-free mice have elongated but thinner villi compared with those found in colonized mice. The longer, thinner shape increases the absorptive surface of the intestine (Figure 1) but at the same time makes the villus more vulnerable and sensitive to infections. Accordingly, colonization of germ-free mice with a normal gut microbiota (conventionalization) causes significant alterations of the villus architecture, which requires increased vascularization in order to oxygenate the villus. Recent results suggest that the increased vascularization is mediated through microbiota-induced angiopoietin-1 expression in the intestinal epithelium.1 Furthermore, germ-free mice have poorly developed gut-associated lymphoid tissue compared with conventionally raised mice. Colonization of germ-free mice with a normal gut microbiota causes a massive infiltration of immune cells to the small intestinal mucosa.2,3,4
Colonization is also associated with the induction of several innate immune responses such as expression of inducible nitric oxide synthase (iNOS, also known as NOS2) and antimicrobial peptides that may be important for protecting the mucosa from potentially harmful bacteria. Conserved microbially associated molecular pattern molecules are recognized by pattern recognition receptors on host cells such as Toll-like receptors and nucleotide-binding oligomerization domain-like receptors that induce a proinflammatory signaling upon microbial recognition. Surprisingly, induction of transcriptional responses in the small intestine did not require functional Toll-like receptor signaling.5 However, signaling through Toll-like receptors is sufficient to alter the gut microbial composition, which has profound effects on host physiology,6–8 although the exact modulators of the microbial ecology remain unidentified. The underlying reasons for these phenotypic differences need to be explored further.
GUT MICROBIOTA AND NUTRITION
While the body is well suited to absorb dietary glucose and to metabolize starch, it does not digest complex fibers, which reach the distal gut and provide energy to the gut microbiota. Accordingly, the gut microbiota efficiently extracts energy from the otherwise indigestible polysaccharides by anaerobic fermentation, which explains why germ-free rats cannot metabolize complex polysaccharides such as cellulose.9 Experiments in obese mice have demonstrated that the obese microbiota is associated with an increased capacity to harvest energy from a polysaccharide-rich diet.10,11 The capacity of the gut microbiota to process different carbohydrates as well as other macronutrients may, at least in part, explain how the gut microbiota can contribute to the development of obesity and its associated diseases. Since nutrients are processed differently by distinct bacterial strains, it is tempting to speculate that only certain combinations of macronutrients and microbiota give rise to disease in genetically predisposed hosts. One such example is the microbial metabolism of phosphatidylcholine by the gut microbiota, which leads to elevated levels of trimethylamines, associated with increased cardiovascular risk.12 Thus, it will become imperative to take diet, host genotype, and the metagenome into consideration in upcoming studies to identify whether the gut microbiota contributes to metabolic disease in humans (Figure 2).
In addition to promoting fermentation of carbohydrates to short-chain fatty acids (SCFAs), the gut microbiota promotes glucose absorption from the small intestine by a yet-unidentified mechanism.13 The increased levels of circulating SCFAs and glucose may act as substrates for de novo lipogenesis in the liver. Previously, it was demonstrated that conventionalization of germ-free mice is associated with increased expression of the key lipogenic transcription factors sterol regulatory element binding protein 1c (SREBP-1c) and carbohydrate element binding protein (ChREBP) as well as expression of the rate-limiting lipogenic enzymes fatty acid synthase and acetyl-CoA caboxylase.13 A recent lipidomic approach confirmed these findings and identified approximately 100 triglyceride species that were elevated in the livers of conventionally raised mice.14 The increased lipogenesis in colonized mice is associated with increased production of hepatic very-low-density lipoprotein, which is transported to the adipose tissue. There, the triglycerides are hydrolyzed by lipoprotein lipase for storage in the adipose tissue.
GUT MICROBIOTA AND OBESITY
The gut microbiota not only affects host metabolism by regulating energy harvest from the gut, it can also regulate host metabolism directly. This became evident when germ-free and conventionalized mice were fed a typical Western diet, which is rich in fat and sucrose. Whereas colonized mice rapidly gained weight, the germ-free mice did not gain more weight than control mice fed a low-fat diet.15 These findings have now been corroborated by several groups. Importantly, Fleissner et al.16 observed that the germ-free mice were only resistant to obesity when sucrose was present in the diet, which further emphasizes the importance of considering the microbiota-diet interaction (Figure 2).
One mechanism by which the gut microbiota regulates development of obesity in the presence of a high-fat, high-sugar diet is by suppressing intestinal expression of angiopoietin-like protein 4 (Angptl4, also known as fasting-induced adipose factor). Angptl4 is a potent inhibitor of lipoprotein lipase and is essential for the microbiota-induced triglyceride storage in the adipose tissue. Furthermore, Angptl4 also promotes fatty acid oxidation in both white adipose tissue and skeletal muscle.15,17 The increased Angptl4 levels in germ-free mice also correlate with increased fatty acid oxidation. Rederivation of Angptl4-/- mice as germ-free suggested that Angptl4 may underlie the protection from diet-induced obesity in germ-free mice.15 Furthermore, germ-free mice have increased activation of AMP-activated kinase (AMPK) in colon, liver, and skeletal muscle.15,18 AMPK is induced when energy is low and promotes catabolic processes such as beta-oxidation while inhibiting lipogenesis. The contribution of AMPK activation in mediating the lean phenotype of germ-free mice remains to be clarified.
Removing the gut microbiota by antibiotics also confers protection from diet-induced obesity as well as reduced body weight gain in genetically obese ob/ob mice.19,20 The protection against obesity is associated with reduced gut permeability and circulating lipopolysaccharide levels, which are associated with obesity. The improved metabolic features in the absence of bacteria were associated with reduced inflammation, suggesting that, in addition to promoting adiposity, the gut microbiota may also contribute to metabolic dysregulation by elevating the inflammatory tone. Furthermore, antibiotic treatment resulted in improved glucose tolerance, similar to that observed in germ-free mice.19,20 Taken together, these results suggest that reduced numbers of bacteria, either in germ-free mice or after antibiotic treatment, are associated with improved metabolic function.
MODIFICATION OF THE GUT MICROBIOTA THROUGH DIETARY MANIPULATION
The gut microbiota can be altered upon feeding of specific fibers that promote “the selective stimulation of growth and/or activity(ies) of one or a limited number of microbial genus(era)/species in the gut microbiota that confer(s) health benefits to the host.”21 Dietary fructans can be used as energy substrates by bacteria such as Bifidobacterium spp. and thus can cause a selective increase in numbers of such bacteria. Dietary supplementation with inulin-type fructans (ITF) increases Bifidobacterium spp. levels and is correlated with reductions in fat mass, glucose intolerance, and circulating lipopolysaccharide levels in both genetic and diet-induced models of obesity.22 ITF supplementation also prevents the overexpression of several host targets related to adiposity and inflammation. The underlying mechanisms for the reduced inflammatory tone remains to be identified, but recent findings suggest that the endocannabinoid system and glucagon-like peptide 2 (GLP-2) are both potent modulators of gut permeability.23,24
GLP-2 is produced upon cleavage of proglucagon, expressed by enteroendocrine L-cells in the distal small intestine and colon. Feeding rodents with ITF increases the number of L-cells and promotes secretion of the active forms of GLP-2 as well as glucagon-like peptide 1 (GLP-1) into the portal vein.25–28 In contrast to the GLP-2-mediated effects on gut permeability, GLP-1 plays a role in prebiotic-driven decreases in appetite, fat mass, and hepatic insulin resistance.28 Interestingly, in healthy volunteers, ITF prebiotics increased the postprandial release of gut peptides (GLP-1 and gastric inhibitory peptide), which were associated with increased satiety, reduced calorie intake, and improved postprandial glycemia.29,30
In addition to serving as an energy source, SCFAs may also act as signaling molecules that activate GPR41 (also known as free fatty-acid receptor 3 [FFAR3]), and recent data suggest that colonization of germ-free mice with Bacteroides thetaiotaomicron and Methanobrevibacter smithii increases production of SCFAs and promotes PYY expression in the distal ileum in a Gpr41-dependent fashion.31 Accordingly, the ability of the gut microbiota to ferment carbohydrates to SCFAs increases the energy harvest. Moreover, the SCFAs resulting from the fermentation of carbohydrates may also function as signaling molecules that can alter the functional capacity of enteroendocrine cells.
Since the identification in 2004 of the gut microbiota as an environmental factor that contributes to adiposity, there has been a rapid development of studies aimed at clarifying the role of the gut microbiota in disease progression. Despite many studies aimed at delineating how the gut microbiota is altered in obesity, there is little consensus, partly due to differences in study design and methodology used (as reviewed elsewhere in the present supplement). These shortcomings may be circumvented by large-scale shotgun metagenomic approaches that do not require polymerase chain reaction amplification of well-phenotyped individuals. These studies reveal not only which bacteria are present but also which genes are associated with obesity or other diseases. Unfortunately, results from such studies will only identify bacteria, or bacterial genes, that are associated with disease or health. In order to investigate causal relationships and disease-causing mechanisms, these metagenomic studies need to be complemented by gnotobiotic animal studies. Recent results demonstrated that germ-free mice can be “humanized” by colonization with resuspended feces from human donors,32 which paves the way for personalized gnotobiotics in which mice are colonized with an individual's microbiota and can be subjected to dietary interventions. By performing similar experiments in germ-free genetically manipulated mice, the underlying signaling pathways contributing to disease development may be identified.
The author gratefully acknowledges the assistance of Gunnel Östergren-Lundén in the preparation of Figure 1.
Funding. Work in the author's laboratory is supported by the Human Frontier of Science Program, the Swedish Research Council, the Swedish Foundation for Strategic Research, the EU-funded ETHERPATHS projects (FP7-KBBE-222639, http://www.etherpaths.org), TORNADO (FP7-KBBE-222720, http://www.fp7tornado.eu/), Åke Wiberg, Torsten and Ragnar Söderberg, the Novo Nordisk Foundations, and an LUA-ALF grant from Västra Götalandsregionen.
Declaration of interest. The author has no relevant interests to declare.