Prospects Overviews J. Alcock et al. 48] and to the hypothalamus [67]. Germ-free mice were also shown to have lower levels of leptin, cholecystokinin, and other satiety 0 sometime Microbes can analogs of behavior nges N-glycated in the gut. B. to increase ly for and host mic of obesity. In wed that mice plasma producing produce GABA activ anti-anxie A for manipu treatment mixture of Lac hormones AgRP (agouti re n the microbiota Bioessays 36: 940-949, 2014 The Autl 943that react to the presence of particular bacteria [48] and to bacterial metabolites such as short-chain fatty acids. Evidence suggests that the vagus nerve regulates eating behavior and body weight. For example, blockade or transection of the vagus nerve has been reported to cause drastic weight loss [49, 50]. On the other hand, vagus nerve activity appears to drive excessive eating behavior in satiated rats when they are stimulated by norepinephrine [51]. These results suggest that gut microbes that produce adrenergic neurochemicals (discussed below) may contribute to over￾eating via mechanisms involving vagal nerve activity. Together these results suggest that microbes have opportunities to manipulate vagus nerve traffic in order to control host eating. Intriguingly, many practices that are known to enhance parasympathetic outflow from the vagus nerve, e.g. exercise, yoga, and meditation, are also thought to strengthen willpower [52] and improve accuracy of food intake relative to energy expenditure [53]. However, increased vagus activity is not always associated with health. One study linked parasympathetic vagus activity with weight loss in patients with anorexia nervosa [54], suggesting that vagus nerve signaling is important in regulating body weight, and sometimes can lead to pathological anorexia. Microbes can influence hosts through hormones Microbes produce a variety of neurochemicals that are exact analogs of mammalian hormones involved in mood and behavior [8, 55–57]. More than 50% of the dopamine and the vast majority of the body’s serotonin have an intestinal source [58, 59]. Many transient and persistent inhabitants of the gut, including Escherichia coli, [8, 55, 56] Bacillus cereus, B. mycoides, B. subtilis, Proteus vulgaris, Serratia marcescens, and Staphylococcus aureus [60] have been shown to manufacture dopamine. Concentrations of dopamine in culture of these bacteria were reported to be 10–100 times higher than the typical concentration in human blood [60]. B. subtilis appears to secrete both dopamine and norepinephrine into their environment, where it interacts with mammalian cells. Transplant of the microbiome from a male to an immature female mouse significantly and stably increases testosterone levels in the recipient [61]. In turn, host enzymes are known to degrade neurotransmitters of bacterial origin. For instance, mammals use monoamine oxidase to silence exogenous signaling molecules, among other functions [62, 63]. This may be evidence for selection on hosts to counteract microbial interference with host signaling. Certain probiotic strains alter the plasma levels of other neurochemicals. B. infantis 35624 raises tryptophan levels in plasma, a precursor to serotonin [64]. The lactic acid producing bacteria found in breast milk and yogurt also produce the neurochemicals histamine [65] and GABA [66]. GABA activates the same neuroreceptors that are targeted by anti-anxiety drugs such as valium and other benzodiazepines. Appetite-regulating hormones are another potential avenue for manipulation of mammalian eating behavior. In mice, treatment with VSL#3, a dietary supplement consisting of a mixture of Lactobacillus strains, reduced hunger-inducing hormones AgRP (agouti related protein) and neuropeptide Y in the hypothalamus [67]. Germ-free mice were also shown to have lower levels of leptin, cholecystokinin, and other satiety peptides [41], hormones that control hunger and food intake partly by affecting vagus nerve signaling. Numerous commen￾sal and pathogenic bacteria manufacture peptides that are strikingly similar to leptin, ghrelin, peptide YY, neuropeptide Y, mammalian hormones that regulate satiety and hunger [68]. Moreover, humans and other mammals produce antibodies directed against these microbial peptides, a phenomenon that could have evolved as a mammalian counter-adaptation to microbial manipulation. Anti-hormone antibody production may be important in maintaining the fidelity of host signaling systems. However, these antibodies also act as auto-antibodies against mammalian hormones [68]. This autoimmune response implies that microbes have the capacity to manipulate human eating behavior (i) directly with peptide mimics of satiety regulating hormones, or (ii) indirectly by stimulating produc￾tion of auto-antibodies that interfere with appetite regulation. The antibody response to microbial analogs of human hormones supports the hypothesis that conflict between host and microbiota influences the regulation of eating behavior. Mucin foraging bacteria control their nutrient supply Several commensal bacteria are known to induce their hosts to provide their preferred nutrients through direct manipulation of intestinal cells. For example, Bacteroides thetaiotaomicron is found on host mucus, where it scavenges N-glycated oligosaccharides secreted by goblet cells in the gut. B. thetaiotaomicron induces its mammalian host to increase goblet cell secretion of glycated carbohydrates [69, 70]. Investigators have shown that another mucin-feeding species, A. muciniphila, also increases the number of mucus producing goblet cells when inoculated in to mice [71]. On the other hand Faecalibacterium prausnitzii, a non-mucus-degrading bacte￾rium that is co-associated with B. thetaiotaomicron, inhibits mucus production by goblet cells [70]. These species provide a proof of principle that gut bacteria can control their nutrient delivery, involving a mechanism that is energetically costly for the host [72]. Intestinal microbiota can affect obesity Evolutionary conflict between the gut microbiome and host may be an important contributor to the epidemic of obesity. In a landmark paper, Backhed and colleagues showed that mice genetically predisposed to obesity remained lean when they were raised without microbiota [73]. These germfree mice were transformed into obese mice when fed a fecal pellet from a conventionally raised obese mouse [74]. Inoculation of germ￾free mice with microbiota from an obese human produced similar results [75]. Mice lacking the toll-like receptor TLR5 became obese and developed altered gut microbiota, hyper￾phagia, insulin resistance, and pro-inflammatory gene expres￾sion [76]. Fecal pellets from these TLR5 knockout mice, when fed to wild type mice, induced the same phenotype. The gut microbes of obese humans are less diverse than the microbiota ....Prospects & Overviews J. Alcock et al. Bioessays 36: 940–949,
2014 The Authors. Bioessays published by WILEY Periodicals, Inc. 943 Review essays