Obesity, Asthma, and the Microbiome

Abstract

Obesity is a risk factor for asthma, but standard asthma drugs have reduced efficacy in the obese. Obesity alters the gastrointestinal microbial community structure. This change in structure contributes to some obesity-related conditions and also could be contributing to obesity-related asthma. Although currently unexplored, obesity may also be altering lung microbiota. Understanding the role of microbiota in obesity-related asthma could lead to novel treatments for these patients.

Obesity is a major public health problem. Within the United States, current estimates from the Center for Disease Control indicate that approximately one-third of the adult population is obese, and another third is overweight. It is well established that obesity is a risk factor for several diseases, including Type 2 diabetes, hypertension, and atherosclerosis. Recent evidence indicates that obesity is also a risk factor for asthma, although the mechanistic basis for this relationship remains ill defined. As described below, it is conceivable that obesity-related changes in the gut and/or lung microbiota contribute to the increased risk of asthma in the obese.

Obesity Is a Risk Factor for Asthma

Both the prevalence and incidence of asthma are increased in the obese (119, 126, 128). Furthermore, in obese asthmatics, weight loss causes substantial reductions in asthma symptoms and reduces airway hyperresponsiveness (AHR) (16, 32). Asthma control is difficult to achieve in many obese asthmatics (33, 107), and steroids are less effective in obese than in lean asthmatics, perhaps because obese asthmatics have a phenotype that is not responsive to steroids (129). Indeed, non-atopic asthma is more common in obese than in lean individuals (139, 142, 148).

Obese mice exhibit innate AHR, a characteristic feature of asthma (4, 73, 79, 118, 120). Innate AHR is observed in genetically obese mice (ob/ob, db/db, and Cpe^fat mice) and in mice with diet-induced obesity (DIO) caused by high-fat feeding. These obesity-related changes in airway responsiveness are observed regardless of the bronchoconstricting agonist used (84), similar to the nonspecific AHR of asthma. Human obesity also augments the decrements in pulmonary function that occur following exposure to the air pollutant ozone (O₃), an asthma trigger (2, 13). Similar effects are observed in mice; compared with lean mice, obese mice develop greater inflammation and greater increases in pulmonary mechanics and in airway responsiveness after acute O₃ exposure (65-67, 84, 121, 122, 151).

Obesity Alters Microbiota in the Gastrointestinal Tract

The human gastrointestinal (GI) tract is home to more than 100 trillion bacteria encompassing more than a thousand different species, many of which are not found outside their hosts (36, 85). Most of these bacteria reside in the colon. The collective genome of these bacteria (the microbiome) has capacities that humans lack. For example, gut bacteria express enzymes that permit metabolism of otherwise indigestible polysaccharides and dietary starches, leading to production of short-chain fatty acids (SCFA), such as butyrate, propionate, and acetate (125, 132), that can ultimately be used for host ATP synthesis or synthesis of other energy substrates in the liver. The gut microbiota also contribute to the synthesis of micronutrients including vitamin K and certain B vitamins, and the absorption of key minerals, especially iron (43, 52, 71). Gut microbiota convert primary to secondary bile acids, participate in xenobiotic detoxification, and have a key role in intestinal epithelial function (55, 64, 71, 83, 99, 124, 134).

The distal gut microbiota are altered in both human and murine obesity (80–82, 108, 135–137, 141). Compared with lean human subjects, obese subjects have less diverse bacterial communities (78, 136). Compared with lean mice, both genetically obese mice and mice with dietary obesity have a greater ratio of bacteria of the Firmicutes to Bacteroidetes phyla (82, 137). These two phyla encompass ∼90% of the bacteria in the normal gastrointestinal tract. Consistent with this observation, the ratio of Firmicutes to Bacteroidetes declines with weight loss in humans (FIGURE 1) (35, 82, 112). Similarly, when feces collected from human subjects after switching to a high-fat, high-sugar diet are transplanted into germ-free mice, the ratio of Firmicutes phylum in germ-free mice increases (138). Nevertheless, the observed association between obesity and a higher Firmicutes-to-Bacteroidetes ratio is not without exception. Several studies have shown either no association or even an opposite finding (41, 59a, 80, 114). Instead, it may not be the most abundant phyla in the gut or their taxonomic composition that is important for health but rather the functional capacities of these bacteria. Closely related taxa can have highly varied functions, whereas distantly related taxa are sometimes grouped together at the functional level (136).

FIGURE 1.

Bacteroidetes and Firmicutes in obese individuals

Before diet, obese individuals had fewer Bacteroidetes and more Firmicutes phyla. After 12, 26, and 52 wk on low-calorie diet, obese individuals increased the relative abundance of Bacteroidetes and decreased the relative abundance of Firmicutes. N = 11–12 per time point. Reproduced from Ref. 82 with permission from Nature Publishing Group.

There are several consequences for the host of obesity-related changes in bacterial populations. For example, the obese microbiome appears to contribute to the adipose tissue inflammation of obesity and to other obesity-related conditions, including Type 2 diabetes and nonalcoholic fatty liver disease (NAFLD) (18, 25, 34, 40, 49, 70, 74, 75, 80). In mice, 4 wk of high-fat diet increases the proportion of bacteria containing lipopolysaccharide (LPS) in the gut (17). Obese mice have increased intestinal permeability, allowing this elevated LPS to leak into the circulation. Consistent with these observations, changes in the gut microbiota induced by antibiotics reverse the insulin resistance and endotoxemia caused by high-fat feeding (18), suggesting that this systemic endotoxemia leads to insulin resistance. Remarkably, even in human subjects, transfer of gut microbiota from nondiabetic into diabetic individuals results in improved insulin sensitivity that persists for several weeks (143). In addition, various probiotic treatments impact both the systemic inflammation and insulin resistance associated with obesity in mice (19, 20).

Changes in the gut microbiota are not only caused by but also promote obesity: gnotobiotic/germ-free mice lacking bacteria do not develop obesity when placed on high-fat diets but do become obese after bacterial reconstitution (7, 135). Similarly, gavage of fecal contents of conventional obese mice into germ-free mice causes the germ-free mice to gain weight over time compared with germ-free mice that received fecal contents from conventional lean mice (137). Four weeks of oral LPS treatment increases the body weight of mice to the same extent as 4 wk of high-fat diet (17). Colonizing germ-free mice with bacteroides thetaiotaomicron leads to greater fermentation of substrates and subsequent production of polysaccharides, increasing adiposity in germ-free rodents (111). Fiaf, an angiopoietin-like protein 4, is repressed by the gut flora and increases fat mass due increased lipoprotein lipase activity (5). Increases in Fiaf also lead to increased expression of genes that regulate mitochondrial oxidation of fatty acids (8). Obese individuals have less circulating SCFAs (114), which have been shown to prevent and reverse high-fat-diet-induced metabolic syndrome (30). These data suggest that obese microbiota also have greater capacity to produce SCFAs that can be used for ATP synthesis by the host.

The microbiome also interacts with the immune system (61, 63, 124) in ways that may alter host metabolism. For example, repopulating the gut of germ-free mice with the microbiota from mice raised in conventional specific pathogen-free (SPF) facilities results in marked changes in gene expression within the intestinal mucosa (38). Mice lacking toll-like receptor 5 (TLR5) have increased ratios of Proteobacteria and Enterobacteria in their gastrointestinal tracts, and develop hallmark features of metabolic syndrome, including low-grade inflammation, hyperlipidemia, insulin resistance, and increased adiposity (141). Furthermore, germ-free mice that are reconstituted with cecal contents of TLR5-deficient mice display many features of the metabolic syndrome (141).

Microbes can also impact host eating behavior. Microbes have been shown to induce dysphoria, a state of unease or dissatisfaction, which may impact feeding behavior (1). The histone deacetylase inhibiting SCFA, butyrate, has been shown to impact the central nervous system and to have an antidepressant-like benefit in mice (113). Several mucin-foraging bacteria, including Bacteroides thetaiotaomicron and A. muciniphila, increase mucus-producing goblet cells, manipulating the host to alter nutrient delivery (117). Together, these studies suggest that microbial ecology may influence food preference, satiety, and behavior.

A Role for Gut Microbiota in Obesity-Related Asthma?

While the impact of gut microbiota on obesity-related asthma has not been examined, there is evidence that gut microbiota have the capacity to impact other asthma phenotypes, including allergic asthma. For example, lungs of mice raised in germ-free conditions have augmented allergic airways responses compared with mice raised in conventional SPF facilities, possibly because of developmental alterations in their immune systems (51, 103). Compared with allergic conventional mice, allergic germ-free mice have elevated eosinophils and lymphocyte infiltration in the airways that is reversed when germ-free mice are recolonized with SPF microbiota (50). However, transfer of microbiota from conventional to germ-free mice imparts immunological protection only when the recolonization occurs early in life (103), perhaps before the immune system is fully developed. An impact of early changes in gut microbiota on adiposity and lung health also has been observed in other studies (23, 93, 100). Young mice given sub-therapeutic doses of antibiotics display increased adiposity and alterations of lipid metabolism (23). Furthermore, the use of perinatal antibiotics in humans has been shown to significantly impact the gut microbiota, leading to alterations in the adaptive immune response and an increased susceptibility to lung inflammatory diseases (109). In humans, both prenatal and postnatal use of antibiotics is associated with an increased risk of developing asthma (93).

How might gut microbiota affect asthma? The gut microbiota have broad effects on the intestinal immune system, including both immuno-stimulatory and immune-regulatory effects (21, 63, 71, 124). Antigen presentation at one mucosal site can drive migration of innate immune cells to other mucosal sites, giving rise to the idea of a common mucosal immune system (27, 76). Through this shared mucosal communication, it is possible that changes in gastrointestinal immune system development may account for the observed differences in the susceptibility of germ-free vs. conventionally raised mice to allergic airways disease. For example, invariant natural killer T cells accumulate in both the colonic lamina propria and the lungs of germ-free vs. conventional mice, at least in part via a CXCL16-dependent mechanism (103).

The gut microbiota could also impact the lungs through alterations in immuno-modulatory cells, such as Th17 cells. Mice treated with antibiotics and germ-free mice have reduced numbers of intestinal IL-17A⁺ T cells (68, 69). Given the importance of IL-17A for neutrophil recruitment (94), such cells may be particularly important for neutrophilic forms of asthma. The gut microbiome also impacts systemic immunity: colonization of germ-free mice with segmented filamentous bacteria increases Th17 cells in the intestines, and also in the central nervous system (62, 69), and gut microbiota also contribute to immune effects in the joints, the pancreas, and the lungs (51, 61, 69, 103). The ability of gut microbiota to affect IL-17A may be especially relevant for obesity-related asthma. Serum IL-17A is elevated in obesity (127), and both BAL IL-17A and lung IL-17A mRNA expression are elevated in obese vs. lean mice (89, 150). Sputum IL-17A is also elevated in obese vs. lean asthmatics (87). Importantly, the innate AHR typically observed in obese mice is absent in mice deficient in IL-17A (FIGURE 2) (73). Given the ability of gut microbiota to perturb intestinal IL-17A⁺ cells and the likelihood of a common mucosal immune system (see above), these data suggest that obesity-related alterations in gut microbiota could play a role in obesity-related asthma via changes in IL-17A. Notably, dietary supplementation with probiotics reduces high-fat-diet-induced increases in gut, liver, and adipose tissue expression of IL-17A (97).

FIGURE 2.

AHR in high-fat diet vs. control diet

Wild-type (WT) mice on high-fat diet (HFD) develop greater AHR compared with WT mice on control chow. Il17a^−/− mice on HFD do not develop a greater AHR than WT or Il17a^−/− mice on control chow. Reproduced from Ref. 73 with permission from Nature Publishing Group.

Small metabolites generated by gut microbiota or metabolites whose generation requires a cascade of enzymes, some of bacterial origin (bacterial mammalian co-metabolites), may also impact the lungs in such a way as to promote asthmatic responses. Such metabolites can diffuse from the gut into the circulation, affecting other target organs, including the lung. Studies using germ-free mice and studies using antibiotic treatment have established that microbiota affect the metabolomes of the intestines, urine, liver, brain, and kidney (9, 24, 54, 55, 86, 88, 90–92). For example, alterations in gut microbiota impact serum levels of many bacterial mammalian co-metabolites, including SCFA, pipecolate, choline, phenol sulfate, and hippurate (88, 133, 149). Microbial-related changes in circulating metabolites have already been correlated with several disease processes. Trimethylamine N-oxide (TMAO), a microbial-dependent metabolite derived from dietary choline, is elevated in the serum of subjects with high atherosclerosis and cardiovascular disease risk (12, 145). Consistent with these findings, transplantation of cecal contents from atherosclerosis-prone but not wild-type strains of mice, transmits enhanced choline-diet-induced atherosclerosis and TMAO levels to antibiotic-treated mice (47). In addition, commensal microbe-derived butyrate induces differentiation of colonic T regulatory cells, which ameliorates T-cell-dependent experimental colitis (42). The importance of these bacterial mammalian co-metabolites extends even to the host's brain function. Four weeks of drinking fermented milk products with probiotics (containing Bifidobacterium animalis subsp Lactis, Streptococcus thermophiles, Lactobacillus bulgaricus, and Lactococcus lactis subsp Lactis) modulates how human subjects perform on attention tasks as well as their resting brain activity (131). Furthermore, serum concentrations of the bacterial-dependent metabolite 4-ethylphenylsulfate (4EPS) are markedly elevated in mouse models of autism and are restored to control levels (along with improvements in autism-like behavior) after treatment with B. fragilis in the food (56).

Microbiota-related changes in circulating concentrations of SCFAs also impact allergic airways responses. Mice given a high-fiber diet have increased gut Bifidobacterium, a bacterium that can ferment dietary fiber to form acetate and propionate, and increased circulating concentrations of these SCFAs (133). In addition, both high-fiber diets and exogenously administered propionate cause alterations in bone marrow hematopoiesis and generation of macrophage and dendritic cell precursors (133). High-fiber diets and propionate also protect against house dust mite-induced allergic inflammation in the lung in a manner that is dependent on GPR41, a receptor for SCFA. Consistent with these observations, low-fiber-fed mice have decreased circulating levels of SCFAs and increased allergic airway response (FIGURE 3) (133). It is conceivable that the altered gut microbiome of obesity may also affect metabolites that circulate to the lungs and affect airway function.

FIGURE 3.

AHR after sensitization and challenge with ovalbumin

Naive mice were neither sensitized nor challenged with ovalbumin. Mice fed a control diet (4% fiber) had moderate AHR after sensitization and challenge with ovalbumin, whereas mice fed a low-fiber diet (<0.3% fiber) had greater AHR after sensitization and challenge with ovalbumin. Reproduced from Ref. 133 with permission from Nature Publishing Group.

Microbiota can also impact metabolism in ways that may impact asthma. For example, gut microbiota can modulate diurnal secretion of glucocorticoids (60). Specifically, lack of microbial signaling leads to overproduction of corticosterone and subsequent development of hypertriglyceridemia, hyperglycemia, and insulin resistance (60). Corticosteroids inhibit the transcription of several inflammatory cytokines and chemokines implicated in asthma and are widely used as asthma therapy (10, 123). However, long-term use of oral glucocorticoids is also associated with increased cardiovascular and metabolic disease risk (140, 147). Hence, altering the microbiota may provide a safer option of controlling asthma symptoms.

The metabolomes of the liver, serum, urine, and adipose tissue are altered in obesity, both in humans and in rodents (26, 48, 72, 96, 104, 115, 116, 144, 152, 155), but whether such changes are the result of obesity-related changes in gut microbiota remains to be established. Insulin resistance is common in obesity, so it is not surprising that glucose, lactate, glycerol, fatty acids, and β-hydroxybutyrate are increased in the blood of obese vs. normal mice (26, 116). Alterations in branched-chain amino acid metabolites are also consistently observed in metabolomic profiles of the blood in obesity (11, 98, 116). Such small metabolites can diffuse across pulmonary capillaries. Receptors for many fatty acids (GPR40, GPR41, GPR49, GPR84, and GPR120), lactate (HCA₁/GPR81), and β-hydroxybutyrate (HCA₂/GPR109A) exist (28, 102). Consequently, obesity-related changes in these moieties, perhaps resulting from alterations in gut microbiota, could impact lung function via direct activation of these receptors.

Obesity, Asthma, and Lung Microbiota

It is now apparent that, although previously considered sterile, the lungs of normal human subjects also possess a microbiome (31, 57, 95). Culture-independent methods of amplifying and sequencing genetic materials unique to prokaryotes (16s rRNA) have enabled us to identify a distinct microbial community in the subglottic airways in some healthy individuals and in the lower airways of patients with obstructive lung diseases, including asthma (22, 39, 44, 53, 58). Indeed, some studies have implicated airway colonization with bacteria in asthma development and severity. For example, detection of Moraxella catarrhalis, Haemophilus influenzae, or Streptococcus pneumoniae in the oropharynx of infants is associated with a significant increase in the risk ratios for recurrent wheeze and childhood asthma (14). Mycoplasma pneumoniae or Chlamydia pneumoniae are found in higher portions in bronchial biopsy specimens of asthmatic patients vs. nonasthmatic controls, and treatment with clarithromycin improves lung function in these patients (77). However, conflicting data also exist. In a study of mild to moderate asthma that was suboptimally controlled with inhaled corticosteroids, Sutherland et al. found that clarithromycin treatment does not improve lung function or airway inflammation (130). Furthermore, neonatal infection of the lungs with Haemophilus pylori actually protects mice against asthma-like features including AHR, lung tissue inflammation, and goblet cell metaplasia, and this protection is abrogated with antibiotic treatment (3).

Such conflicting results may be at least in part related to the phenotypic and mechanistic variability of asthma. Asthmatic patients with greater airway microbial diversity also have greater airway hyperresponsiveness (59). In addition, subjects with greater improvements in AHR with clarithromycin treatment also have higher baseline bacterial diversity (59). In addition, the sputum of treatment-resistant severe asthmatics contains relatively higher abundances of Moraxella catarrhalis, Haemophilus influenzae, and Streptococcus pneumoniae that correlate with decreased lung function, elevated BAL neutrophils, and IL-8 concentrations (46).

The establishment of lung microbial communities is a dynamic process with contributions from early life practices, environmental exposures, and the host's lung-specific immune development. Changes in the lung microbiome at a critical window of childhood development are associated with alterations in airway response to allergen exposure (44). Children in countries with higher incidence of breast feeding, children who grow up with farm animals or pets, and children who consume nonpasteurized milk or contaminated water often have lower rates of allergy and asthma (37, 101, 105, 146). Whether other asthma triggers such as ozone and particulate air pollution can also affect lung microbiota remains to be established.

It is conceivable that obesity also alters lung microbial communities. The hyperlipidemia and hyperglycemia that often accompany obesity are likely to result in increased lipid and glucose concentrations within the lung lining fluid as well. Changes in these nutrients might create environments that permit some bacteria to flourish. The airway closure that frequently accompanies obesity (110) will also create hypoxic niches that affect bacterial survival. A study of oral microbiota shows BMI-related differences in diversity and abundances of salivary bacteria, and the presence of one species (Selenomonas noxia) can separate 98.4% of overweight women from the lean (45).

Conclusions

High throughput sequencing techniques indicate that microbiota are present in virtually every anatomical site in the human organism and that dysregulation of these microbiota plays an important role in the progression of multiple diseases. Obesity alters the microbiota in the gastrointestinal tract. Such changes could play a role in obesity-related asthma via alterations in production of metabolites with effects in the lung, such as SCFAs, by contributing to the systemic inflammation of obesity via changes in the host metabolism, insulin sensitivity, and even feeding behavior, and by changes in the immune system, especially IL-17A⁺ immune cells. It is also conceivable that obesity alters the lung microenvironment in such a way as to alter lung microbiota that impact asthma development and/or therapeutic responses (FIGURE 4).

FIGURE 4.

Schematic representation of ways in which altered gut (and possibly lung) microbiota in obesity might lead to asthma

The hypothesis that obesity-related alterations in gut and/or lung microbiota contributes to obesity-related asthma has several important therapeutic implications. It may be that host immune system development can be shaped by factors affecting the developing gut microbial communities, including early life choices and exposure to environmental bacteria. It is also feasible that altering the gut microbiota could ameliorate obesity-related asthma. Probiotic and prebiotic treatments have been shown to be effective against other obesity-related conditions (29, 106, 153). Transplantation of microbiota from healthy to diseased individuals is increasingly used for treatment of some intestinal illnesses, especially Clostridium difficile infection (15, 154), and even has been shown to increase insulin sensitivity in patients with metabolic syndrome (143). Understanding the role of gut and airway microbiota in obesity-related asthma could pave the way for development of microbiota-based treatments for this difficult-to-treat group.