Role of the Commensal Microbiota in Normal and Pathogenic Host Immune Responses

Cell Host & Microbe

Review Role of the Commensal Microbiota in Normal and Pathogenic Host Immune Responses Dan R. Littman1,2,* and Eric G. Pamer3,4,* 1Molecular

Pathogenesis Program Hughes Medical Institute The Kimmel Center for Biology and Medicine of the Skirball Institute, New York University School of Medicine, New York, NY 10016, USA 3Immunology Program, Lucille Castori Center for Microbes, Inflammation and Cancer, Sloan-Kettering Institute 4Infectious Diseases Service, Department of Medicine Memorial Sloan-Kettering Cancer Center, New York, NY 10065, USA *Correspondence: [email protected] (D.R.L.), [email protected] (E.G.P.) DOI 10.1016/j.chom.2011.10.004 2Howard

The commensal microbiota that inhabit different parts of the gastrointestinal (GI) tract have been shaped by coevolution with the host species. The symbiotic relationship of the hundreds of microbial species with the host requires a tuned response that prevents host damage, e.g., inflammation, while tolerating the presence of the potentially beneficial microbes. Recent studies have begun to shed light on immunological processes that participate in maintenance of homeostasis with the microbiota and on how disturbance of host immunity or the microbial ecosystem can result in disease-provoking dysbiosis. Our growing appreciation of this delicate host-microbe relationship promises to influence our understanding of inflammatory diseases and infection by microbial pathogens and to provide new therapeutic opportunities. Introduction The mammalian gut harbors an enormous diversity of commensal microbes that far outnumber the cells of the host. The composition of the microbial community is thought to reflect the coevolution of host and microbes to achieve a balanced, mutually beneficial state. The microbiota contribute to host nutrition and energy balance and to the development or maintenance of a robust immune system. In turn, the host provides a niche in which the individual microbes can persist within their community, from which they may obtain essential nutrients, ensuring their transmission and retention within the host species. Although it has long been proposed that the microbiota have important roles in a range of systemic diseases, only the advent of new experimental tools, including gnotobiotic animals colonized with defined commensal microbes, mouse inflammation models, and culture-independent approaches, particularly massive parallel sequencing technology, has resulted in new insights into how microbiota interact with the host immune system. Recent studies have provided firm evidence that skewing of the commensal community, often referred to as ‘‘dysbiosis,’’ can result in inflammatory diseases not only of the intestine, but also of organs at distal sites. Such diseases can be triggered not only by pathogenic microbes, but also by otherwise harmless commensal microbes or those that are normally held in check by the microbial ecosystem and/or the metabolic state and immune response of the host. Thus, disturbance of this homeostasis by intrinsic or extrinsic influences, e.g., treatment with broad-spectrum antibiotics, can result in life-threatening dysbiosis. Deep sequencing of the fecal metagenome of healthy humans has demonstrated remarkable interindividual variability of intestinal microbial populations, but it has been suggested that a limited number of balanced symbiotic states, referred to as enterotypes, become established in the gut (Arumugam et al.,

2011; Eckburg et al., 2005; Suau et al., 1999). Whether these different enterotypes influence health and disease susceptibility remains unclear. However, in light of recently described associations between specific intestinal commensal bacteria, T cell differentiation, and inflammatory bowel disease (IBD), it seems very likely that the complex constellation of commensal and symbiotic bacteria in the gut influences the intestine’s innate immune tone and modulates resistance to infection or susceptibility to inflammatory disease. Thus, after nearly 150 years of adhering to Koch’s postulates and investigating interactions between specific bacterial pathogens and their human hosts, microbiologists and immunologists now confront the more complex challenge of dissecting the impact of a network of nonpathogenic microbes on the mucosal immune system. In this review, we will attempt to break down the variables that contribute to homeostasis of microbiota and components of host immunity. As will become apparent, it remains difficult to make generalizations, as animal models reveal multiple scenarios that result in inflammatory disease. What has become clear, however, is that the combination of the host genotype and the composition of the microbiota are the major determinants of disease. The influence of the environment—e.g., diet, toxin, or antibiotic exposure that can contribute to pathological inflammatory conditions, perhaps triggering dysbiosis and/or a breakdown in epithelial barrier function—will also be discussed. Components of the Intestinal Immune System Organized Lymphoid Tissues The mucosal tissues in the GI tract are considered to harbor more cells of the immune system than all the secondary lymphatic tissues combined. The composition and organization of cells in the GI tract are still being defined, particularly with the emergence of novel markers that allow further subdivision of lineages or states of differentiation. In addition to the Peyer’s Cell Host & Microbe 10, October 20, 2011 ª2011 Elsevier Inc. 311

Cell Host & Microbe

Review

Figure 1. Schematic of Intestinal Immune System Cells Discussed in This Review Cryptopatches consist of RORgt+ lymphoid tissue inducer cells (that comprise one of the subsets of ILCs) and dendritic cells (whose precise phenotype has not been determined). Signals transmitted from microbiota to cryptopatches result in influx of B lymphocytes to form isolated lymphoid follicles (not shown in the figure). Induced Treg cells control the expansion and effector activity of Th1 and Th17 cells as well as ILCs. TCRgd cells are not shown, but also contribute to inflammatory cytokine production.

patches and mesenteric lymph nodes, which are secondary lymphoid tissues that develop during fetal life, there are a series of tertiary lymphoid tissue structures that develop postnatally, which include isolated lymphoid follicles (ILFs) and cryptopatches in the small intestine and colonic patches in the large intestine (van de Pavert and Mebius, 2010). Development of all such lymphoid tissues requires lymphotoxin-mediated signals transmitted by innate lymphoid cells (ILCs) known as lymphoid tissue inducer (LTi) cells. LTi cells and tertiary lymphoid tissues are absent in mice deficient for the expression of the transcription factors Id2 or RORgt. Cryptopatches consist of RORgt+ LTi cells surrounded by CD11c+ dendritic cells and underlie the single cell layer of the intestinal epithelium (Figure 1). They receive NOD1-dependent signals from the microbiota and require signals mediated by the LTi cells for recruitment of B lymphocytes and formation of ILFs (Bouskra et al., 2008; Tsuji et al., 2008). ILFs are sites of mucosal secretory immunoglobulin A (sIgA) production, which in the intestine occurs through both T cell-dependent and -independent mechanisms and requires microbial colonization (Macpherson et al., 2008). Lymphoid Cells In addition to the organized tissues, there are multiple other immune cell subsets represented throughout the intestinal lamina propria (LP) and the epithelium. These include a variety of ILCs and myeloid cells as well as multiple types of T lymphocytes that express either the ab or the gd T cell antigen receptor (TCR) and exhibit an activated or memory-like phenotype. The LP is the site most enriched in differentiated ab CD4+ T cells, particularly Th17 cells and regulatory T cells (Treg). Th17 cells produce IL-17A, IL-17F, and IL-22, cytokines that have important roles in clearing microbes that reach the LP and in maintaining the mucosal barrier, thus preventing inflammation induced by microbial products that breach the epithelium. When not properly regulated, these cells can be potent mediators of inflammation. Treg cells, which express the transcription factor Foxp3 and 312 Cell Host & Microbe 10, October 20, 2011 ª2011 Elsevier Inc.

often produce the anti-inflammatory cytokine IL-10, are most prominent in the large intestine, the site of the highest concentration of microbiota. Many of the Foxp3+ cells acquire their regulatory phenotype in the LP microenvironment and are designated induced regulatory T cells (iTreg) to distinguish them from cells that acquire Foxp3 during thymic development (designated natural, or nTreg) (Curotto de Lafaille and Lafaille, 2009). ILCs include several phenotypic variants that mirror T helper cells in their cytokine profiles. Cells that express very high levels of IL-22 and/or lower levels of IL-17 make up a large fraction of ILCs, and like Th17 cells, they are distinguished by the expression of the transcription factor RORgt and by IL-23R and the chemokine receptor CCR6. LTi cells fall in this category, although it is unclear if they can be classified based on a distinct phenotype. IL-23 produced by myeloid cells is required for the differentiation of LTi cells, particularly for their expression of IL-22, and the interaction of LTi expressed CCR6 with its ligand, CCL20, is thought to be critical for their localization. IFNgproducing NK cells and recently described lymphoid cells that produce IL-4 or IL-13 are also present in the intestinal LP and can be modulated in the presence of diverse microbes (reviewed in Spits and Di Santo, 2011). Myeloid Cells Among myeloid cells, CD103+ DC and CD103 CX3CR1hi DC (also designated as macrophages because they are thought to be nonmigratory) have received the most attention, in part because the former were shown to be endowed with the enzymatic machinery to produce retinoic acid, which enhances the conversion of CD4+ T cells to the iTreg phenotype. The myeloid cells, particularly the monocyte-derived cells that express high levels of the fractalkine receptor CX3CR1, extend processes through the tight junctions of intestinal epithelial cells and make contact with lumenal contents (Figure 1). Extension of dendrites is dependent on signals from lumenal commensal bacteria by way of the Toll-like receptor (TLR) adaptor MyD88 in epithelial cells (Chieppa et al., 2006) and on expression of CX3CR1 and its ligand (Niess et al., 2005). Sampling of lumenal contents by a CX3CR1-dependent mechanism has been reported as important for immunological tolerance to orally ingested antigen, but a role for this process in immune responses to microbial products has not yet been established (Hadis et al., 2011). Influence of the Microbiota on the Intestinal Immune System Colonization of the intestinal tract with diverse microbes has a profound influence on the development and function of both innate and adaptive branches of the immune system. In antibiotic-treated mice, the reduction in MyD88-mediated innate immune signals induced by bacterial products renders the mucosa more permeable to commensal bacteria and less capable of postinjury repair (Vaishnava et al., 2008). Induction of TLR signaling, e.g., by supplying LPS, can substitute for gut bacteria in protecting the epithelium from potential damage (Rakoff-Nahoum et al., 2004). Defects in various other innate immunity signaling modules, e.g., NOD2 or components of inflammasome signaling pathways involved in release of IL-1b and IL-18, can result in increased risk for IBD or in uncontrolled growth of invasive enteropathic bacteria in humans and in

Cell Host & Microbe

Review mouse models. Thus, the diverse pattern recognition receptors do not discriminate between molecular components of normally harmless commensals and pathogenic invasive bacteria. How the host immune system selectively clears pathogens while tolerating the presence of commensals remains one of the key unanswered questions in microbe-host mutualism. One possibility is that commensals, by virtue of being noninvasive, do not gain access to immune system cells with innate receptors. However, as discussed below, there is now ample evidence that commensal bacteria do indeed influence the differentiation state of immune system cells in the intestinal LP. Moreover, innate signaling receptors, e.g., various TLRs, are present on the surface of epithelial cells, albeit not always at the apical surface or readily accessible through the mucus layer that is rarely penetrated by commensals. An alternative possibility, which we favor, is that coevolution of host and commensal microbial communities has led to a mutualistic de´tente that provides a means of avoiding inflammation while preventing elimination of key microbial constituents through the activation of innate and adaptive immune components. This homeostasis is likely also promoted by microbe-induced Tregs that restrain potential inflammatory responses by both adaptive and innate arms of the immune system. In this setting, a disturbance of homeostasis can result in aberrant activation of the largely promiscuous innate receptors, in subsequent tissue damage, and potentially in systemic inflammation. A major shift in the composition of the commensal community can also result in outgrowth of invasive pathogenic microbes, as illustrated by infections following treatment with antibiotics and discussed in detail below. Effector T Cell Responses The proportion of Th17 and iTreg cells differs according to the site in the gut (e.g., small versus large intestine) and the composition of the microbiota. The latter varies widely between different institutional animal facilities, and this factor needs to be considered when interpreting results reported for various models of inflammation. Mice lacking expression of TCRa develop a spontaneous colitis that is dependent on the presence of commensal microbiota (Bhan et al., 1999). This was an early observation that suggested a loss of regulation of inflammatory processes in the absence of ab T cells. Mutations resulting in loss of IL-2 or its receptor subunits similarly resulted in microbiota-dependent spontaneous colitis, which can now be ascribed to the absence of Treg cells in these mice (Sadlack et al., 1993). Adoptive transfer of naive T cells into B and T cell-deficient mice results in colitis only if mice are colonized with intestinal microbiota, which contribute to the expansion of effector T cells in the absence of Treg cells in the LP (Maloy and Powrie, 2011). Colitis in this commonly used T cell transfer model requires expression in the transferred cells of RORgt and the receptor for IL-23, which are both essential for the differentiation of Th17 cells. IFNgproducing Th1 cells also have a prominent role in this and related models of colitis, but Th17 cells are essential for disease pathogenesis, and fate-mapping studies have shown that Th1 cells in inflammatory settings can be derived from IL-17A-expressing precursors (Hirota et al., 2011). Microbial communities, which include hundreds of bacterial species as well as fungi and viruses that have received considerably less attention, influence the abundance of ab and gd T cells

and the production of sIgA by plasma cells in the intestinal mucosa. Much recent attention has focused on how the microbiota govern the balance of CD4+ T cell subsets in the intestinal LP. Our current understanding of how this is achieved is limited to a handful of individual bacterial species that can direct the accumulation of either Th17 cells or Treg cells in the LP of small and large intestine. Gnotobiotic mice, which lack the microbial flora, have no Th17 cells in either small or large intestine. Colonization of such germ-free mice with the Gram-positive anaerobe segmented filamentous bacterium (SFB) results in accumulation of Th17 cells, primarily in the small intestine, where SFB associate closely with the intestinal epithelial cells (GaboriauRouthiau et al., 2009; Ivanov et al., 2009). SFB influence the balance of intestinal CD4+ T cells not only in monoassociated mice, but also in mice that harbor a diverse microbiota. For example, C57BL/6 mice with a diverse microbiota but lacking SFB have few intestinal Th17 cells, but they have at least five times as many Th17 cells after they are colonized with SFB or exposed to feces from mice that have SFB in their commensal community (Ivanov et al., 2008). SFB represent only one of what is likely to be a large number of microbes that elicit Th17 cells. A variety of bacteria generally considered pathogenic, e.g., Citrobacter rodentium, Klebsiella pneumoniae (Happel et al., 2005), and Salmonella typhimurium (Godinez et al., 2008, 2009; Schulz et al., 2008), as well as fungi, e.g., Candida albicans (Conti et al., 2009), induce Th17 cells in the intestine or at other mucosal surfaces. The cytokines produced by the Th17 cells are essential for protecting the mucosal barrier from these pathogenic microbes. In accord with this, colonization of mice with SFB provides protection from subsequent colonization with C. rodentium. Although the protection can be ascribed, at least in part, to induction of Th17 cells, it cannot be ruled out that SFB have additional beneficial properties that limit access of pathogenic bacteria to the mucosal epithelium. For example, short-chain fatty acids produced by commensal bacteria provide substantial intestinal barrier protection (Maslowski et al., 2009), and SFB may similarly employ this metabolite to reduce colonization with pathogenic bacteria. Polarization of antigen-stimulated T cells toward the Th17 profile requires the presence in the microenvironment of IL-1b, IL-6, IL-23, and TGF-b. Fungi produce b-glucans, whose interaction with Dectin-1 selectively induces production of IL-23 by dendritic cells (Reid et al., 2009). Pathogenic bacteria presumably gain access to the epithelium and penetrate it, activating multiple innate sensors, including TLRs and NLRs, resulting in production of the Th17-inducing cytokines. How SFB or other commensals promote Th17 cell differentiation is less clear. SFB have the unique property of adhering closely to intestinal epithelial cells, particularly in the terminal ileum. To do so, SFB must penetrate the thick layer of mucus that overlays the epithelium. Study of the biology of SFB has been hampered by the inability thus far to culture this spore-forming anaerobe. However, the recent elucidation of the SFB genomic sequence promises to provide new tools to elucidate its mechanism of action (Kuwahara et al., 2011; Prakash et al., 2011; Sczesnak et al., 2011). For example, SFB encode multiple glycan foraging enzymes, which may facilitate their ability to penetrate mucin, and several flagellins, among which at least three can activate TLR5. Whether TLR signaling is involved in SFB-induced Th17 Cell Host & Microbe 10, October 20, 2011 ª2011 Elsevier Inc. 313

Cell Host & Microbe

Review cell differentiation is not yet known. However, induction of Th17 cells by transfer of SFB-containing microbiota into SFB-deficient mice does not require MyD88 or TLR3 and hence appears to be independent of TLR engagement (Ivanov et al., 2008). The ability of SFB to gain close access to the epithelium and, potentially, to processes extended by dendritic cells may underlie their capacity to induce Th17 cell differentiation, but characterization of the signaling pathways involved awaits further investigation. Induction of Th17 cytokines requires ligation of the TCR, but the specificity of LP Th17 cells does not appear to be restricted to microbial antigen. Mice with monospecific T cells have intestinal Th17 cells and monoassociation of mice with SFB results in as many Th17 cells as in mice with a broad spectrum of intestinal microbes (Ivanov et al., 2009; Lochner et al., 2011). Th17 cells may therefore be in large part self-reactive T cells circulating through the LP, where a reduced threshold for TCR signaling and the presence of appropriate cytokines results in their polarization. Such T cells may then circulate out of the intestine and would be available for recruitment to other sites in the body, where inappropriate activation can result in organ-specific Th17 cell-mediated autoimmunity (Lee et al., 2011; Wu et al., 2010; Mathis and Benoist, 2011). A large fraction of gd T cells in the LP also express the cytokines associated with Th17 cells. Unlike ab T cells, gd T cells are induced to produce effector cytokines by IL-23 and IL-1b even in the absence of TCR ligation (Sutton et al., 2009). In this manner, they resemble ILCs and are thought to provide early protection against pathogenic microbes (Jensen et al., 2008). The number of gd T cells has not been shown to be affected by commensal microbiota, although intraepithelial lymphocytes (IELs) bearing the ab TCR do expand upon colonization of germ-free mice with SFB (Umesaki et al., 1995). IELs may provide a critical link between the microbiota and innate intestinal responses, as the transcriptional program of gd IEL is regulated by commensals and their ablation increases gut permeability to commensals following injury (Ismail et al., 2011). ILCs in Inflammation There has been considerable recent interest in the functions of ILCs that are particularly abundant in the intestinal LP. Intestinal ILCs that produce IL-17 and/or IL-22 are dependent on RORgt for their differentiation and appear to participate in a variety of inflammatory processes. In a mouse model of innate immune system-mediated IBD, elicited by colonization of RAG-deficient mice with Helicobacter hepaticus, IL-23R and RORgt were required for the development of Thy-1+CD127+ ILCs that mediated the disease through production of IL-17 and IFNg (Buonocore et al., 2010). Polymorphisms in IL-23R are strongly associated with incidence of Crohn’s disease, consistent with the importance of these effector cells along with T cells in the pathogenesis of IBD in humans. Similar to effector T lymphocytes, the inflammatory function of ILCs is also subject to suppression by Treg cells. In contrast to the proinflammatory function of ILCs revealed in the innate IBD model, a protective function of ILCs was demonstrated in response to infection with enteropathic C. rodentium. Following infection, IL-22 production was far higher in ILCs than in T cells. IL-22, which regulates expression of antimicrobial peptides by intestinal epithelial cells, is essential in protecting the host from infection with C. rodentium (Zheng et al., 2008). 314 Cell Host & Microbe 10, October 20, 2011 ª2011 Elsevier Inc.

Production of IL-22 by the ILCs was dependent on the presence of IL-23 and was sufficient to protect RAG-deficient mice from lethal infection. Mice reconstituted with T cells and depleted of ILCs with anti-isotype antibody succumbed to infection, suggesting that ILCs are more critical than T cells for barrier protection in this model (Sonnenberg et al., 2011). A role for microbiota in the accumulation of mucosal ILCs has been suggested, but results have not been consistent, and it is also unclear if there are discrete effects on specific subsets of ILCs. Colonization of mice with SFB resulted in increased LP Th17 cells, but no obvious change in ILCs (Ivanov et al., 2009). Because SFB reduced the amount of C. rodentium associated with colonic epithelium, it will be important to investigate beneficial functions of SFB that are independent of Th17 cell induction. Treg Cells and the Microbiota Recent studies indicate that Treg cells are also regulated by the microbiota. In conventional animal facilities, representation of Treg cells among CD4+ T cells is much higher in the large intestine and, to a lesser degree, in the small intestine than in lymph nodes or other secondary lymphoid organs. Removal of microbial flora results in an increased fraction of Treg cells in the small intestine, although there are fewer total CD4+ T cells and Th17 cells are absent. In the large intestine, in contrast, there is a substantial reduction in Treg cell numbers in germ-free mice. This reduction is largely in the iTreg population, which is defined by the absence of expression of the transcription factor Helios. The iTreg cells were largely restored by colonization of germfree mice with a mixture of 46 Clostridial strains, enriched in representation of clusters IV and XIVa (Atarashi et al., 2011). Colonization with the mixed Clostridia was also sufficient to induce IL-10, an anti-inflammatory cytokine, and CTLA4, a molecule that inhibits T cell activation, in a substantial fraction of the iTreg cells. When neonates kept in a conventional specific pathogen-free facility were gavaged with the Clostridia, the animals were protected from inflammation induced with disodium sulfate (DSS) or oxazalone and displayed lower levels of ovalbumininduced IgE. Moreover, Clostridia elevated the proportion of Treg cells not only in the gut LP, but also in other organs, including lung, liver, and skin. Properties of Clostridia that contribute to induction of iTreg cells and the specificity of the iTreg cells in this setting are not yet known. However, a notable recent study has shown that iTreg cells have a unique TCR repertoire that confers specificity for commensal bacteria, and other results suggest that bacterial products can influence the Th17:Treg balance, potentially by favoring iTreg cell differentiation or expansion. The specificity of LP T cells was addressed recently by Lathrop et al. (Lathrop et al., 2011), who showed that intestinal iTreg cells have a TCR repertoire distinct from that of naive and effector LP CD4+ T cells. TCRs of most iTreg cells were found to recognize antigens present in the heat-killed lumenal content from mice colonized with commensal microbiota, but not from germ-free mice, and some of the TCRs were reactive with single bacterial species. When the TCRs of microbiota-reactive Treg cells were introduced into naive T cells, those newly activated cells expanded and became colitogenic upon transfer into RAGdeficient mice. These results suggest that exposure of naive T cells to antigens derived from intestinal microbiota induces the differentiation and expansion of iTreg cells rather than

Cell Host & Microbe

Review effector CD4+ T cells, thus enforcing immune tolerance to individual bacterial species. Because these experiments focused on restricted pairs of TCRa and b chains, it remains unclear whether other combinations of TCRs may bias differentiation toward effector T cells. How exposure to microbial antigen favors iTreg cell differentiation versus expansion of effector cells or polarization to a particular effector phenotype (e.g., Th17 cells) is an important question for future investigation. The answer may be found in microbial products that induce production of distinct cytokines by cells in the microenvironment or even by the T cells themselves. Thus, polysaccharide A (PSA) produced by Bacteroides fragilis, a commensal bacterium of humans, protects mice from colitis induced upon colonization with Helicobacter hepaticus (Mazmanian et al., 2008). The mechanism of protection has been proposed to involve TLR2dependent induction in T cells of IL-10, which restrains proinflammatory Th17 cells (Round et al., 2011). In contrast to the Clostridia, PSA+ B. fragilis had only a modest influence on the proportion of intestinal Treg cells, and it is not known if PSA influences iTreg cell abundance in vivo. Nevertheless, it is likely that there are multiple commensal bacteria that have the capacity to influence the differentiation of iTreg cells, potentially through regulating the levels of TGF-b, retinoic acid, and IL-2, which contribute to iTreg cell differentiation. The molecular mechanisms by which bacterial products affect release of these and other mediators will no doubt be the subject of intense investigation. Together, the recent results raise the exciting possibility that specific commensals or their products may be therapeutically beneficial in a variety of systemic or organ-specific autoimmune diseases. Inflammatory Bowel Disease and Dysbiosis: Lessons from Experimental Models A series of recent reports have highlighted the importance of a balanced intestinal microbiota for maintenance of a healthy mucosal barrier. Some of the experiments were prompted by the results of genome-wide association studies (GWAS) aimed at detecting polymorphisms associated with IBDs in humans, while others were based on examination of mice with mutations in key genes involved in adaptive and innate immune responses. Animal models based on GWAS results have shed light on the importance of innate signals and an intact autophagy machinery in barrier function, particularly when challenged by viral infection, but there has been only limited analysis of the effects of autophagy on the microbiota (Cadwell et al., 2010). Mice with mutations in many of the genes involved in innate sensing have been examined in various models of IBD, and the role of the intestinal microbiota in such mice is generating much interest. A notable recent study described the consequence of impaired inflammasome signaling on sensitivity of the intestinal mucosa to mild chemical disruption with DSS (Elinav et al., 2011). Inactivation of NLRP6, resulting in reduced IL-18 production, rendered mice highly susceptible to DSS colitis. Remarkably, cohoused wild-type mice became similarly susceptible, and this was ascribed to sharing of a dysbiotic microbiota by mice within the colony. The mutant mice displayed a shift in the fecal microbiota, and this bias was transmitted to cohoused animals. There was a marked increase of a novel member of the Prevotellaceae family, in the phylum Bacteroidetes, and of a member of the

phylum TM7, and antibiotics that reduced association of these bacteria with colonic crypts suppressed the DSS colitis phenotype. Spontaneous colonic inflammation and elevated production by the epithelium of the chemokine CCL5 were observed in both mutant and cohoused wild-type mice, and DSS-induced disease was dependent on CCL5. This is thus an example of a dominant effect of commensal dysbiosis, with apparently asymptomatic inflammation in the setting of an undisturbed gut epithelium. It is conceivable that similarly biased microbial communities exist in humans and contribute to disease only if the mucosal barrier is perturbed, e.g., by toxins or diet. It should be noted that a direct role has not been demonstrated for the bacterial species that prevail in this model, and their expansion may be secondary to the host response to other constituents of the microbiota. Colonization of mice with individual cultured bacterial species will be needed to demonstrate a direct role of such bacteria in colitis. The effector mechanism in this disease model is also yet to be elucidated, and it is thus unclear if T cell function is disrupted, e.g., Th1 cell differentiation in the setting of reduced IL-18. It will also be of great interest to learn the nature of the NLRP6 ligand that contributes to healthy homeostasis. Another example of dysbiosis in the setting of a compromised immune system was described in mice defective for both IL-10 and TGF-b receptor signaling (Bloom et al., 2011). These signaling pathways were implicated by GWAS in human IBD. Mice deficient in IL-10R2 and expressing a dominant-negative TGF-bR2 in T cells developed spontaneous colitis with high penetrance, but disease was prevented by administering metronidazole and ciprofloxacin, antibiotics typically employed in treatment of ulcerative colitis. By gavaging mutant mice after neonatal antibiotic treatment with commensal anaerobes grown in different selective media, it was demonstrated that commensal Bacteroides, specifically B. thetaiotaomicron and B. vulgatus, were sufficient to induce colitis. Unlike the previously described model, wild-type mice cohoused with the mutant animals or gavaged with the same Bacteroides species were resistant to disease. In addition, there was no enrichment of the Bacteroides in mutant versus wild-type mice, although there was enrichment of commensal Enterobacteriaceae, such as E. coli, that did not transmit disease. The effector mechanisms mediating colitis are not yet known, but there was an increase in Th1 cells, as expected in the absence of Treg cells (which require TGF-b signaling for differentiation and/or survival) and of IL-10- and TGF-b-mediated inhibition of IFNg expression. Whether IFNg and/or ILC functions are required for the effector phase of disease will await further studies. This model thus represents a second category of dysbiosis associated with immune dysfunction, in which disruption of a host immune homeostatic process uncovers an inflammatory signal initiated by a defined bacterial species. Unlike the inflammation initiated in the absence of NLRP6 signaling, the microbiota in this case did not appear to have any pathogenic effect in wild-type mice, and it is even unclear if there was indeed dysbiosis that contributed to disease, rather than a loss of host regulatory signaling. A related model of IBD was described in TRUC mice, which are defective for the transcription factor T-bet and lack an adaptive immune system due to RAG2 deficiency (Garrett et al., Cell Host & Microbe 10, October 20, 2011 ª2011 Elsevier Inc. 315

Cell Host & Microbe

Review

Figure 2. Examples of How Commensal or Pathogenic Microbiota Can Influence Immune Homeostasis in the Gut SFB and various clostridia can induce accumulation of Th17 or iTreg, respectively. Commensal bacteria like Prevotella do not associate with the epithelium unless there is dysbiosis, as observed in mice with mutations in NLRP6 or IL18. Such bacteria may elicit antigen-specific iTreg cells, preserving immune tolerance to the microbiota. It is hypothesized that displacement of the bacteria to the epithelial surface results in inflammatory signals and changes in the epithelial permeability. Invasive bacteria like Klebsiella or Citrobacter are typically held in check by the host immune system, but can breach the epithelium when pathways related to Th17 cells or ILCs are compromised.

2007). These mice also developed colitis with a high penetrance, and mutant offspring had significantly different commensal communities compared to wild-type siblings. Disease in this model was due to elevated production of TNF-a by T-bet-deficient dendritic cells and was prevented by antibiotic treatment or blockade of TNF-a as well as by adoptive transfer of Treg cells. The TRUC colitis was communicable to T-bet-sufficient mice and was more severe in RAG2-deficient mice than in wild-type mice, likely due to absence of Treg cells in the former. The TRUC mice, but not RAG2-deficient mice, had culturable K. pneumoniae and Proteus mirabilis in their feces, and transfer of these bacteria to RAG2-deficient mice induced the colitis phenotype. However, germ-free TRUC mice were resistant to disease induction with these bacteria, indicating that an endogenous microbial community was additionally required for colitis to develop. The K. pneumoniae and P. mirabilis were found to penetrate the mucus layer of the colon and to be closely associated with the apical surface of the epithelial cells (Garrett et al., 2010). The mechanism by which they contribute to inflammation is unclear, but it likely involves either direct interactions of microbial products with cells at the epithelial surface or indirect effects conveyed through host immune cells sensing the mislocalized bacteria. This example of a transmissible dysbiosis may also apply to human IBD, but unlike that observed with NLRP6-deficient mice, this disease scenario involves what are commonly thought of as pathogenic bacteria. The microbiota-dependent inflammation described in the three examples above occurs in the setting of diverse genetic defects in innate and adaptive immune functions and may be horizontally transmissible under special circumstances (Figure 2). Pre-existing immunological impairment may not be a 316 Cell Host & Microbe 10, October 20, 2011 ª2011 Elsevier Inc.

prerequisite for dysbiosis-mediated inflammation, however. Treatment with antibiotics may shift the microbiota, allowing for outgrowth of invasive pathogenic bacteria, e.g., vancomycin-resistant enterococci or Clostridium difficile, as described below. It cannot be ruled out that exposure to certain antibiotics, diet, or environmental toxins can also favor dysbiosis in which normally harmless commensals can trigger inflammatory processes. Although such examples have not yet been reported in experimental models, the apparent proinflammatory activity of Prevotella-enriched microbiota in the mucosa of wild-type animals in the Elinav et al. (2011) study discussed above suggests that alterations in the representation of individual commensal constituents may result in disease induction. For example, a community that achieves a stable bias favoring inducers of effector cells (e.g., Th1, Th17, ILCs) over inducers of Treg cells may result in perpetuation of an inflammatory state. The term ‘‘pathobiont’’ has been coined to describe commensal bacteria that can be pathogenic in such a setting (Round and Mazmanian, 2009). While single commensal micro-organisms that induce disease may exist, it will be necessary to consider the impact of the broader community of microbiota in potentiation of disease. Accumulating data from metagenomic analyses of microbiota present at different stages of inflammatory diseases will likely provide critical insight into how microbial communities participate in such diseases. Antibiotic Effects on the Intestinal Microbiota and Immune Defense While colonization of germ-free mice with specific commensal organisms has been informative, this approach can be limited by the dependence of different commensal microbes upon their fellow commensals. An alternative approach to characterizing the impact of commensal bacteria on mucosal immunity is to alter the microbiota by antibiotic administration. Antibiotics differ in terms of antimicrobial spectrum, tissue localization, and excretion, properties that can be also affected by the route of administration. Some intravenously administered antibiotics are excreted by the kidneys and do not achieve significant levels in the GI tract, while other antibiotics are concentrated by the liver in the bile and are excreted into the intestinal lumen. Short-term administration of Clindamycin, an antibiotic that is concentrated in bile and thus achieves high levels in the intestine, dramatically alters the microbial composition of the human colon for up to 2 years (Jernberg et al., 2007), while systemic administration of vancomycin, which is cleared by the kidneys, leaves the intestinal microbiota intact. Current understanding of the effect of specific antibiotics on short- and long-term colonization with diverse bacteria is limited. Administration of ciprofloxacin, a commonly used antibiotic with little activity against obligate anaerobic bacteria, altered the abundance of roughly one-third of intestinal bacteria in three healthy volunteers, with a significant loss of microbial diversity during treatment (Dethlefsen et al., 2008). Although the diversity of the microbiota returned within weeks of antibiotic cessation, there were long-term losses of some bacterial taxa that persisted for 6 months (Dethlefsen et al., 2008). Most studies of the human gut microbiota have focused on fecal samples, largely because access to the small intestine is technically and anatomically challenging. Studies in mice,

Cell Host & Microbe

Review however, have demonstrated that the microbial composition of the small intestine and the colon are distinct, with markedly different responses to antibiotic administration. Treatment of mice with ampicillin/metronidazole/bismuth or with cefoperazone revealed dramatic effects on the intestinal microbiota and also showed that mice quickly share intestinal flora as a result of coprophagia (Antonopoulos et al., 2009). Cohousing antibiotic-treated mice with untreated mice results in the rapid normalization of the microbial flora. Oral streptomycin or vancomycin alters the flora of the murine cecum and colon and results in the expansion of g-proteobacteria populations (Sekirov et al., 2008). Antibiotic treatment also impacts T cell populations in the colonic LP, with decreased IFNg and IL17A production (Hill et al., 2010). Treatment of mice with oral vancomycin or ampicillin markedly reduced bacterial density in the ileum and cecum (Ubeda et al., 2010). Mice treated with oral vancomycin often became heavily colonized with bacteria belonging to the Enterobacteriaceae family, perhaps because this antibiotic has limited activity against Gram-negative bacteria. Treatment with ampicillin, on the other hand, did not result in expansion of Enterobacteriaceae. Upon cessation of antibiotic treatment, bacterial density rapidly returned to pretreatment levels, but the microbial composition remained altered (Ubeda et al., 2010). An interesting aspect of postantibiotic rebound microbial expansion is its variability between similarly treated mice. Thus, when separately housed mice with similar microbial flora were treated with the same antibiotic regimen, there were blooms and crashes of gut bacterial populations, resulting in vastly different patterns at 2–3 weeks after cessation of antibiotic treatment (Ubeda et al., 2010). The consequences of these differences in the microbiota for susceptibility to infection and immune defense remain unclear. Recent studies of the fecal metabolome have demonstrated that broad-spectrum antibiotic administration affects the levels of 87% of detectable host molecules, including steroid hormones, eicosanoids, and bile acids (Antunes and Finlay, 2011; Antunes et al., 2011). It is reasonable to speculate that the marked and currently unpredictable fluctuations in the intestinal microbiota that follow antibiotic treatment result in equally unpredictable effects on immune defenses. As discussed above, intestinal microbes and their products stimulate innate immune receptors and drive expression of antimicrobial factors by intestinal epithelial cells. For example, Nod2-mediated signals drive the expression of some cryptdins, thereby enhancing resistance to intestinal invasion by Listeria monocytogenes (Kobayashi et al., 2005) and limiting the density of commensal bacteria in the ileum (Petnicki-Ocwieja et al., 2009). Commensal bacteria also induce the expression of RegIIIg, a secreted C-type lectin that kills Gram-positive bacteria (Cash et al., 2006). Expression of RegIIIg is dependent on TLR signaling (Brandl et al., 2007), and MyD88-deficient mice, despite colonization by commensal bacteria, do not express RegIIIg. TLR-mediated signaling can occur in cells that are not of hematopoietic origin, and MyD88 expression in Paneth cells is sufficient to drive RegIIIg expression (Vaishnava et al., 2008). Antibiotic administration to mice greatly reduces RegIIIg expression and renders mice susceptible to colonization with antibioticresistant bacteria such as vancomycin-resistant Enterococcus faecium (VRE) and Clostridium difficile (Jarchum et al., 2011; Kinnebrew et al., 2010). Orally administered LPS stimulates

RegIIIg expression in the ileum of mice that have been treated with broad-spectrum antibiotics. In this setting, LPS-induced RegIIIg expression results from MyD88-mediated signaling in nonhematopoietic, presumably epithelial cells, and thus simulates commensal-driven RegIIIg induction (Brandl et al., 2008). In contrast to commensal or oral LPS-driven RegIIIg expression, systemic flagellin administration drives RegIIIg expression that is dependent on TLR-5 or MyD88 expression in cells of hematopoietic origin (Kinnebrew et al., 2010). Systemic flagellin administration also corrects the loss in mucosal innate immune defenses in the ileum and protects against colonization of mice with VRE or lethal infection with Clostridium difficile. Flagellin-induced RegIIIg expression is highly dependent on IL-22 production, although the flagellin-responsive and IL-22-producing cells remain to be identified (Kinnebrew et al., 2010). Interaction between Microbiota and Intestinal Pathogens Bona fide intestinal bacterial pathogens cause disease in humans in the absence of known immunodeficiency or defects in the host’s intestinal microbiota. Thus, organisms such as Shigella flexneri, enteroinvasive Escherichia coli, Salmonella enterica, and Vibrio cholera cause dysentery or profuse diarrhea upon oral ingestion by otherwise healthy individuals. Experimental studies in humans demonstrated that the inoculum of bacteria required for successful infection varies markedly for these pathogens. Shigella flexneri is the most infectious and causes disease in some individuals following ingestion of as few as 100 live bacteria (DuPont et al., 1989). However, while 100 live bacteria caused infection in 39% of volunteers, increasing the dosage to 105–108 only increased the rate of infection to 64%. Thus, although not previously exposed to Shigella, 36% of volunteers remained resistant to inoculation with very high doses of this pathogen (DuPont et al., 1989). In contrast to Shigella infections, successful infection, defined as prolonged high fever and diarrhea, required much higher doses of viable Salmonella typhi, Escherichia coli, or Vibrio cholera. Challenge of human volunteers with 103 live Salmonella typhi did not cause infection, and oral administration of 105, 107, and 108–9 live bacteria caused infection in roughly 40%, 50%, and 95% percent of volunteers (Glynn et al., 1995). In the case of Vibrio cholera, oral inoculation of healthy human volunteers with 104 bacteria did not cause diarrhea, while infection with 106 live bacteria induced diarrhea in 27% and 88% of volunteers, depending on the infecting strain (Cash et al., 1974). Interpretation of these studies is difficult, as patients likely differed in terms of previous exposures and overall immune competence. Nevertheless, it is reasonable to speculate that differences in the intestinal microbiota of these volunteers accounted for marked differences in susceptibility to infection by these GI pathogens. Indeed, extensive animal studies indicate that vulnerability to infection by intestinal pathogens is markedly increased by prior disruption of the intestinal microbiota. In an effort to establish animal models of infection with Vibrio cholera or S. enterica, guinea pigs and mice, respectively, were treated with streptomycin and found to become highly susceptible to infection (Bohnhoff and Miller, 1962; Freter, 1955). Vancomycin and streptomycin treatment rendered mice highly susceptible to S. enterica colitis, while treatment with Cell Host & Microbe 10, October 20, 2011 ª2011 Elsevier Inc. 317

Cell Host & Microbe

Review metronidazole enabled colonization but not colitis. In contrast, metronidazole treatment of the 129S1/SvImJ mouse strain, which had a distinct commensal microbiota, resulted in severe S. enterica colitis. Analysis of the microbiota suggested a correlation between the presence of bacteria belonging to the Porphyromonadaceae family and resistance to S. enterica colitis (Ferreira et al., 2011). Metronidazole treatment of mice has been found to decrease the thickness of the colonic mucus layer, resulting in increased susceptibility to C. rodentium infection (Wlodarska et al., 2011). Obligate anaerobic bacteria of the gut have been proposed to provide colonization resistance against pathogenic organisms such as Salmonella (Bohnhoff et al., 1964a, 1964b). Products of obligate anaerobic metabolism, such as acetate and other short-chain fatty acids, inhibit the growth of Salmonella. Recent studies characterized the ability of different strains of Bifidobacterium to provide protection against enterotoxigenic E. coli infection. Strains of Bifidobacterium that produced acetate enhanced resistance to E. coli infection by restricting epithelial damage and toxin absorption (Fukuda et al., 2011). Based on the recent genome sequencing, SFB are also predicted to have glycan transporters and enzymes to generate acetate (Kuwahara et al., 2011; Prakash et al., 2011; Sczesnak et al., 2011). This, along with enhanced Th17 induction in the LP, may contribute to the ability of SFB to confer resistance against infection with C. rodentium (Ivanov et al., 2009). Host inflammatory responses to infection by intestinal pathogens, such as S. enterica or C. rodentium, alter the composition of the microbiota, leading to a loss of bacterial membership and expansion, in particular, of bacteria belonging to the Enterobacteriaceae family (Lupp et al., 2007; Stecher et al., 2007). S. enterica-induced expansion of Enterobacteriaceae has been correlated with the entry of neutrophils into the intestinal lumen (Sekirov et al., 2010). Inflammation-induced alteration of the microbiota enhances S. enterica persistence and expansion in the intestinal lumen by mechanisms that remain incompletely defined (Stecher et al., 2007). S. enterica, in contrast to most commensal microbes, is capable of benefitting from inflammatory responses by selective lipocalin-2 resistance (Raffatellu et al., 2009) and the ability to utilize tetrathionate as a terminal electron acceptor (Winter et al., 2010). Comparison of mice colonized with a range of bacterial populations demonstrated that colonization with bacteria belonging to the Enterobacteriaceae family rendered mice more susceptible to S. enterica infection (Stecher et al., 2010). This finding suggests that Enterobacteriaceae either condition the gut to be more receptive to other members of this family or the abundance of Enterobacteriaceae reflects a more receptive intestinal environment, perhaps reflecting a paucity of other, inhibitory commensal bacteria. While reintroduction of complex microbial populations eliminated S. enterica from the GI tract and was independent of intestinal host IgA production, which members of the microbiota perform this function remains unclear (Endt et al., 2010). Speculation that obligate anaerobes contribute to the protective function, as initially suggested by the work from Bohnhoff nearly 50 years ago, remains enticing. Whether commensal microbes enlist the host innate immune system to eliminate S. enterica from the gut remains unclear and requires further investigation. 318 Cell Host & Microbe 10, October 20, 2011 ª2011 Elsevier Inc.

Clostridium difficile is a Gram-positive, spore-forming rod that causes a spectrum of intestinal diseases extending from relatively mild diarrhea to toxic megacolon (Kelly and LaMont, 2008; O’Connor et al., 2009). C. difficile-associated disease (CDAD) generally occurs in hospitalized patients or individuals residing in other long-term health care facilities and is almost always associated with the administration of antibiotics. Since 2005, the incidence of CDAD has increased dramatically in the United States (and other countries also), from less than 150,000 cases in 2000 to an estimated 500,000 cases in 2006, with approximately 15,000–20,000 deaths (Rupnik et al., 2009). C. difficile causes disease in the colon by secreting two toxins, referred to as toxin A and toxin B, which enter intestinal epithelial cells and glucosylate Rho GTPases, disrupting the cytoskeleton and destroying the integrity of the mucosal barrier. C. difficile is acquired by the oral ingestion of spores from a contaminated environment (Rupnik et al., 2009). Hospitals are notorious for contamination with C. difficile spores, and C. difficile spores are notoriously difficult to inactivate. The GI tract of humans can be colonized with C. difficile. The rate of C. difficile colonization in the general population is approximately 2%–3%, while colonization rates in hospitalized patients ranges from 10% to 25% (Bartlett and Gerding, 2008; Simor et al., 2002). Although specific strains of C. difficile are associated with more severe disease, it remains unclear which host or microbial factors determine recurrence and the extent or duration of CDAD. A recent study investigating stool samples from patients with and without C. difficile infection suggested that recurrent infections occur in a setting of diminished microbial diversity (Chang et al., 2008). Recent studies demonstrate that antibiotic treatment of C57BL/6 mice renders them susceptible to infection with C. difficile, with mortality correlating with the inoculum of spores (Chen et al., 2008; Lawley et al., 2009). Administration of five antibiotics to C57BL/6 mice obtained from Jackson Laboratory followed by a single dose of clindamycin rendered mice highly susceptible to C. difficile infection, with pathologic changes in the cecum and colon that closely resembled changes seen during human infection (Chen et al., 2008). Intestinal disease and mortality vary markedly in antibiotic-treated mice, depending on the strain of C. difficile and also the host mouse strain. Mice surviving C. difficile colitis developed long-term immunity, although the mechanism of immune protection remains undefined. Mice that did not receive antibiotics became colonized upon inoculation with C. difficile and shed low numbers of spores for several months after inoculation. Treatment of colonized mice with clindamycin markedly increased shedding of spores and correlated with an antibiotic-induced decrease in microbial diversity in fecal samples (Lawley et al., 2009). Although supershedder mice had high frequencies of C. difficile bacteria in stool (roughly 5% of total bacteria), they did not develop overt disease. However, histologic examination of the cecum and colon revealed evidence of mild epithelial inflammation and damage. MyD88-deficient mice were highly susceptible to C. difficile infection and developed severe weight loss and diarrhea (Lawley et al., 2009). It remains unclear whether mortality resulted directly from C. difficile infection or from intestinal epithelial damage and dissemination of other intestinal bacteria. Nod1-deficient mice had decreased neutrophil recruitment to

Cell Host & Microbe

Review the colon and increased bacterial translocation across the epithelium, leading to markedly increased mortality (Hasegawa et al., 2011), suggesting that early neutrophil recruitment to the colon during C. difficile infection creates a barrier that prevents systemic dissemination of intestinal microbes. While our understanding of immune defense against C. difficile is only rudimentary, it is likely that experimental investigation of the newly developed murine model of C. difficile colitis will provide new insights and lead to new approaches to prevent this infection. Intestinal Domination by Commensal Bacteria The ability of the normal intestinal microbiota to limit the integration of exogenously introduced, foreign bacteria into the existing population is referred to as colonization resistance (van der Waaij et al., 1971; Vollaard and Clasener, 1994). Colonization resistance contributes to resistance against infection by intestinal pathogens, which, as a rule, must grow in the intestinal lumen in order to cause disease. Studies of colonization resistance have principally focused on competition between microbes, leading to a search for mechanisms by which one microbe suppresses its competitors. However, recent advances in our understanding of mucosal immune compartments have led to the hypothesis that commensal microbes activate innate immune mechanisms that suppress competing microbes, adding a layer of complexity that is only beginning to be investigated. This new perspective may provide insights into some classic studies that quantified the expansion of some intestinal microbes upon antibiotic administration. For example, early studies demonstrated that antibiotic administration enhanced intestinal colonization by Enterobacter cloacae and K. pneumoniae, two members of the Enterobacteriaceae family, while untreated mice rapidly eliminated these organisms from the GI tract (Vollaard and Clasener, 1994). Studies in humans also demonstrated that antibiotics that eliminated obligate anaerobic bacteria increased the density of Enterobacteriaceae and Enterococcaceae (Vollaard and Clasener, 1994). The progressive development of antibiotic resistance in these bacterial families and their ability to thrive in the GI tract when colonization resistance is compromised has created a significant and costly clinical problem. For example, VRE has, over the past two decades, become a major cause of hospital-acquired bacteremia (Kamboj et al., 2010). Clinical studies demonstrated that antibiotics that kill anaerobic bacteria result in marked increases in the density of VRE in human fecal samples (Donskey et al., 2000). VRE, in both mice and humans, can overwhelmingly colonize the intestinal lumen, at times representing 99% of the bacterial flora (Ubeda et al., 2010). Although VRE, unlike C. difficile, is not known to express virulence factors to enable intestinal epithelial invasion or damage, overly dense colonization of the gut is associated with VRE blood stream infection (Ubeda et al., 2010). While the complex microbiota inhabiting the intestine stimulates expression of antimicrobial factors, such as RegIIIg, it is possible that innate immune defenses are compromised when microbial diversity is lost and one microbial species predominates. Further studies are required to investigate the hypothesis that organisms such as VRE and various Enterobacteriaceae become problematic because they do not stimulate the host’s innate immune defenses.

Figure 3. Factors Involved in Maintenance or Disruption of the Mutualistic Balance of Host Functions and the Microbiota The schematic depicts the homeostasis that exists between microbiota (including both commensals and ‘‘pathobionts’’), components of host immunity, and the host tissue (represented with blue arrows). This homeostasis can be disturbed and dysbiosis introduced (represented by black arrows) by extrinsic factors, including antibiotics, diet, environmental toxins, and acute enteric pathogens, and intrinsic factors such as the host’s genetic makeup. Understanding and manipulating these parameters to reset homeostasis could be exploited as strategies to reverse the deleterious consequences of infection with pathogens or antibiotic-induced dysbiosis.

Strategies to Reverse Antibiotic-Induced Susceptibility to Intestinal Infection and Restore Homeostasis The intestinal microbiota enhance resistance to infection by bacterial pathogens and rogue commensals by multiple mechanisms. First, intestinal commensals, in particular obligate anaerobic bacteria, directly inhibit growth of aerobic bacteria such as Enterococcaceae and Enterobacteriaceae. Second, intestinal commensal bacteria stimulate the expression of antimicrobial factors such as RegIIIg, thereby protecting the epithelial layer from invasion by lumenal bacteria and also regulating the balance of constituents of the microbial community. Third, stimulation of both innate and adaptive immune responses by microbiota results in production of cytokines that strengthen the epithelial barrier (e.g., IL-22) and in recruitment of phagocytes that can rapidly clear any invasion resulting from the breach in barrier functions. A fourth likely mechanism, largely conceptual at this time, is host tolerance to commensal microbes. This mechanism would not involve the immune system, but rather the integrity of the host infrastructure, e.g., metabolic pathways or cellular structures, allowing the host to tolerate high microbial loads. Examples of host tolerance to microbes that could become pathogenic abound in nature, but molecular mechanisms have largely remained unexplored (Schneider and Ayres, 2008). Although the relative contributions of the different mechanisms noted above are currently unknown, they could potentially be exploited in strategies to reverse the deleterious consequences of infection with pathogens or dysbiosis, e.g., following antibiotic-mediated depletion of the intestinal microbiota (Figure 3). Which bacterial taxa mediate colonization resistance remains unclear, but determining their identity may enhance Cell Host & Microbe 10, October 20, 2011 ª2011 Elsevier Inc. 319

Cell Host & Microbe

Review the development of probiotic agents that could be administered with or following antibiotic treatment in order to maintain resistance against microbial pathogens. How intestinal bacteria mediate colonization resistance also remains incompletely defined, but determining the identity of bacterially produced factors that inhibit the growth of pathogens will likely lead to new approaches to restrict their growth and pathogenesis. Stimulation of TLRs or other innate immune receptors to drive the expression of host antimicrobial factors represents another promising approach to enhance mucosal resistance during and following antibiotic treatment. An additional option might be to administer bacteria or their products that stimulate specific T cell subsets (or ILCs) to induce production of protective cytokines, such as IL-22. TLR and other signals could also be bypassed by administering such cytokines to directly induce antimicrobial protein expression by epithelial cells. Finally, a growing understanding of the concept of host tolerance to intestinal microbes may also reveal molecular processes that can be manipulated to enhance the host infrastructure and confer protection even in the setting of dysbiosis. Future Perspectives Although our understanding of the relationship of microbiota and the immune system is rudimentary, recent studies suggest that, by targeting this axis, it will be possible to apply preventive and therapeutic approaches to multiple inflammatory and infectious diseases. For example, insights into how individual microbes or their specific products alter the proportions of effector to Treg cells in the intestine could potentially be applied to therapies for autoimmune disease or to preventing graft rejection. Manipulation of the intestinal immune system could also provide therapies for growing problems such as antibiotic-induced dysbiosis and infections with antibiotic-resistant bacteria. A further understanding of how different branches of the host immune system interact in the face of concurrent microbial challenges may also provide better defenses against viral infections: for example, microbiota have been reported to influence the outcome of airway infection with influenza virus, and the human immunodeficiency virus replicates extensively in the LP, where it preferentially depletes Th17 cells. The intersection of the microbiota, the host genetic makeup, and the host immune system promises to provide exciting new therapeutic opportunities in the near future. ACKNOWLEDGMENTS E.G.P. is supported by NIH grant RO1 AI042135. D.R.L. and E.G.P. are also supported by NIH grant RC2 AR058986. REFERENCES Antonopoulos, D.A., Huse, S.M., Morrison, H.G., Schmidt, T.M., Sogin, M.L., and Young, V.B. (2009). Reproducible community dynamics of the gastrointestinal microbiota following antibiotic perturbation. Infect. Immun. 77, 2367– 2375. Antunes, L.C., and Finlay, B.B. (2011). A comparative analysis of the effect of antibiotic treatment and enteric infection on intestinal homeostasis. Gut Microbes 2, 105–108. , P., Borchers, C.H., and Finlay, B.B. Antunes, L.C., Han, J., Ferreira, R.B., Lolic (2011). Effect of antibiotic treatment on the intestinal metabolome. Antimicrob. Agents Chemother. 55, 1494–1503.

320 Cell Host & Microbe 10, October 20, 2011 ª2011 Elsevier Inc.

Arumugam, M., Raes, J., Pelletier, E., Le Paslier, D., Yamada, T., Mende, D.R., Fernandes, G.R., Tap, J., Bruls, T., Batto, J.M., et al; MetaHIT Consortium. (2011). Enterotypes of the human gut microbiome. Nature 473, 174–180. Atarashi, K., Tanoue, T., Shima, T., Imaoka, A., Kuwahara, T., Momose, Y., Cheng, G., Yamasaki, S., Saito, T., Ohba, Y., et al. (2011). Induction of colonic regulatory T cells by indigenous Clostridium species. Science 331, 337–341. Bartlett, J.G., and Gerding, D.N. (2008). Clinical recognition and diagnosis of Clostridium difficile infection. Clin. Infect. Dis. 46 (Suppl 1 ), S12–S18. Bhan, A.K., Mizoguchi, E., Smith, R.N., and Mizoguchi, A. (1999). Colitis in transgenic and knockout animals as models of human inflammatory bowel disease. Immunol. Rev. 169, 195–207. Bloom, S.M., Bijanki, V.N., Nava, G.M., Sun, L., Malvin, N.P., Donermeyer, D.L., Dunne, W.M., Jr., Allen, P.M., and Stappenbeck, T.S. (2011). Commensal Bacteroides species induce colitis in host-genotype-specific fashion in a mouse model of inflammatory bowel disease. Cell Host Microbe 9, 390–403. Bohnhoff, M., and Miller, C.P. (1962). Enhanced susceptibility to Salmonella infection in streptomycin-treated mice. J. Infect. Dis. 111, 117–127. Bohnhoff, M., Miller, C.P., and Martin, W.R. (1964a). Resistance of the Mouse’s Intestinal Tract to Experimental Salmonella Infection. I. Factors Which Interfere with the Initiation of Infection by Oral Inoculation. J. Exp. Med. 120, 805–816. Bohnhoff, M., Miller, C.P., and Martin, W.R. (1964b). Resistance of the Mouse’s Intestinal Tract to Experimental Salmonella Infection. Ii. Factors Responsible for Its Loss Following Streptomycin Treatment. J. Exp. Med. 120, 817–828. Bouskra, D., Bre´zillon, C., Be´rard, M., Werts, C., Varona, R., Boneca, I.G., and Eberl, G. (2008). Lymphoid tissue genesis induced by commensals through NOD1 regulates intestinal homeostasis. Nature 456, 507–510. Brandl, K., Plitas, G., Schnabl, B., DeMatteo, R.P., and Pamer, E.G. (2007). MyD88-mediated signals induce the bactericidal lectin RegIII gamma and protect mice against intestinal Listeria monocytogenes infection. J. Exp. Med. 204, 1891–1900. Brandl, K., Plitas, G., Mihu, C.N., Ubeda, C., Jia, T., Fleisher, M., Schnabl, B., DeMatteo, R.P., and Pamer, E.G. (2008). Vancomycin-resistant enterococci exploit antibiotic-induced innate immune deficits. Nature 455, 804–807. Buonocore, S., Ahern, P.P., Uhlig, H.H., Ivanov, I.I., Littman, D.R., Maloy, K.J., and Powrie, F. (2010). Innate lymphoid cells drive interleukin-23-dependent innate intestinal pathology. Nature 464, 1371–1375. Cadwell, K., Patel, K.K., Maloney, N.S., Liu, T.C., Ng, A.C., Storer, C.E., Head, R.D., Xavier, R., Stappenbeck, T.S., and Virgin, H.W. (2010). Virus-plussusceptibility gene interaction determines Crohn’s disease gene Atg16L1 phenotypes in intestine. Cell 141, 1135–1145. Cash, R.A., Music, S.I., Libonati, J.P., Snyder, M.J., Wenzel, R.P., and Hornick, R.B. (1974). Response of man to infection with Vibrio cholerae. I. Clinical, serologic, and bacteriologic responses to a known inoculum. J. Infect. Dis. 129, 45–52. Cash, H.L., Whitham, C.V., Behrendt, C.L., and Hooper, L.V. (2006). Symbiotic bacteria direct expression of an intestinal bactericidal lectin. Science 313, 1126–1130. Chang, J.Y., Antonopoulos, D.A., Kalra, A., Tonelli, A., Khalife, W.T., Schmidt, T.M., and Young, V.B. (2008). Decreased diversity of the fecal Microbiome in recurrent Clostridium difficile-associated diarrhea. J. Infect. Dis. 197, 435–438. Chen, X., Katchar, K., Goldsmith, J.D., Nanthakumar, N., Cheknis, A., Gerding, D.N., and Kelly, C.P. (2008). A mouse model of Clostridium difficile-associated disease. Gastroenterology 135, 1984–1992. Chieppa, M., Rescigno, M., Huang, A.Y., and Germain, R.N. (2006). Dynamic imaging of dendritic cell extension into the small bowel lumen in response to epithelial cell TLR engagement. J. Exp. Med. 203, 2841–2852. Conti, H.R., Shen, F., Nayyar, N., Stocum, E., Sun, J.N., Lindemann, M.J., Ho, A.W., Hai, J.H., Yu, J.J., Jung, J.W., et al. (2009). Th17 cells and IL-17 receptor signaling are essential for mucosal host defense against oral candidiasis. J. Exp. Med. 206, 299–311.

Cell Host & Microbe

Review Curotto de Lafaille, M.A., and Lafaille, J.J. (2009). Natural and adaptive foxp3+ regulatory T cells: more of the same or a division of labor? Immunity 30, 626–635. Dethlefsen, L., Huse, S., Sogin, M.L., and Relman, D.A. (2008). The pervasive effects of an antibiotic on the human gut microbiota, as revealed by deep 16S rRNA sequencing. PLoS Biol. 6, e280. Donskey, C.J., Chowdhry, T.K., Hecker, M.T., Hoyen, C.K., Hanrahan, J.A., Hujer, A.M., Hutton-Thomas, R.A., Whalen, C.C., Bonomo, R.A., and Rice, L.B. (2000). Effect of antibiotic therapy on the density of vancomycin-resistant enterococci in the stool of colonized patients. N. Engl. J. Med. 343, 1925– 1932. DuPont, H.L., Levine, M.M., Hornick, R.B., and Formal, S.B. (1989). Inoculum size in shigellosis and implications for expected mode of transmission. J. Infect. Dis. 159, 1126–1128. Eckburg, P.B., Bik, E.M., Bernstein, C.N., Purdom, E., Dethlefsen, L., Sargent, M., Gill, S.R., Nelson, K.E., and Relman, D.A. (2005). Diversity of the human intestinal microbial flora. Science 308, 1635–1638. Elinav, E., Strowig, T., Kau, A.L., Henao-Mejia, J., Thaiss, C.A., Booth, C.J., Peaper, D.R., Bertin, J., Eisenbarth, S.C., Gordon, J.I., and Flavell, R.A. (2011). NLRP6 inflammasome regulates colonic microbial ecology and risk for colitis. Cell 145, 745–757. Endt, K., Stecher, B., Chaffron, S., Slack, E., Tchitchek, N., Benecke, A., Van Maele, L., Sirard, J.C., Mueller, A.J., Heikenwalder, M., et al. (2010). The microbiota mediates pathogen clearance from the gut lumen after non-typhoidal Salmonella diarrhea. PLoS Pathog. 6, e1001097. Ferreira, R.B., Gill, N., Willing, B.P., Antunes, L.C., Russell, S.L., Croxen, M.A., and Finlay, B.B. (2011). The intestinal microbiota plays a role in Salmonellainduced colitis independent of pathogen colonization. PLoS ONE 6, e20338. Freter, R. (1955). The fatal enteric cholera infection in the guinea pig, achieved by inhibition of normal enteric flora. J. Infect. Dis. 97, 57–65. Fukuda, S., Toh, H., Hase, K., Oshima, K., Nakanishi, Y., Yoshimura, K., Tobe, T., Clarke, J.M., Topping, D.L., Suzuki, T., et al. (2011). Bifidobacteria can protect from enteropathogenic infection through production of acetate. Nature 469, 543–547. Gaboriau-Routhiau, V., Rakotobe, S., Le´cuyer, E., Mulder, I., Lan, A., Bridonneau, C., Rochet, V., Pisi, A., De Paepe, M., Brandi, G., et al. (2009). The key role of segmented filamentous bacteria in the coordinated maturation of gut helper T cell responses. Immunity 31, 677–689.

Hasegawa, M., Yamazaki, T., Kamada, N., Tawaratsumida, K., Kim, Y.G., Nu´n˜ez, G., and Inohara, N. (2011). Nucleotide-binding oligomerization domain 1 mediates recognition of Clostridium difficile and induces neutrophil recruitment and protection against the pathogen. J. Immunol. 186, 4872– 4880. Hill, D.A., Hoffmann, C., Abt, M.C., Du, Y., Kobuley, D., Kirn, T.J., Bushman, F.D., and Artis, D. (2010). Metagenomic analyses reveal antibiotic-induced temporal and spatial changes in intestinal microbiota with associated alterations in immune cell homeostasis. Mucosal Immunol. 3, 148–158. Hirota, K., Duarte, J.H., Veldhoen, M., Hornsby, E., Li, Y., Cua, D.J., Ahlfors, H., Wilhelm, C., Tolaini, M., Menzel, U., et al. (2011). Fate mapping of IL-17producing T cells in inflammatory responses. Nat. Immunol. 12, 255–263. Ismail, A.S., Severson, K.M., Vaishnava, S., Behrendt, C.L., Yu, X., Benjamin, J.L., Ruhn, K.A., Hou, B., DeFranco, A.L., Yarovinsky, F., and Hooper, L.V. (2011). Gammadelta intraepithelial lymphocytes are essential mediators of host-microbial homeostasis at the intestinal mucosal surface. Proc. Natl. Acad. Sci. USA 108, 8743–8748. Ivanov, I.I., Frutos, Rde.L., Manel, N., Yoshinaga, K., Rifkin, D.B., Sartor, R.B., Finlay, B.B., and Littman, D.R. (2008). Specific microbiota direct the differentiation of IL-17-producing T-helper cells in the mucosa of the small intestine. Cell Host Microbe 4, 337–349. Ivanov, I.I., Atarashi, K., Manel, N., Brodie, E.L., Shima, T., Karaoz, U., Wei, D., Goldfarb, K.C., Santee, C.A., Lynch, S.V., et al. (2009). Induction of intestinal Th17 cells by segmented filamentous bacteria. Cell 139, 485–498. Jarchum, I., Liu, M., Lipuma, L., and Pamer, E.G. (2011). Toll-like receptor 5 stimulation protects mice from acute Clostridium difficile colitis. Infect. Immun. 79, 1498–1503. Jensen, K.D., Su, X., Shin, S., Li, L., Youssef, S., Yamasaki, S., Steinman, L., Saito, T., Locksley, R.M., Davis, M.M., et al. (2008). Thymic selection determines gammadelta T cell effector fate: antigen-naive cells make interleukin-17 and antigen-experienced cells make interferon gamma. Immunity 29, 90–100. Jernberg, C., Lo¨fmark, S., Edlund, C., and Jansson, J.K. (2007). Long-term ecological impacts of antibiotic administration on the human intestinal microbiota. ISME J. 1, 56–66. Kamboj, M., Chung, D., Seo, S.K., Pamer, E.G., Sepkowitz, K.A., Jakubowski, A.A., and Papanicolaou, G. (2010). The changing epidemiology of vancomycin-resistant Enterococcus (VRE) bacteremia in allogeneic hematopoietic stem cell transplant (HSCT) recipients. Biol. Blood Marrow Transplant. 16, 1576–1581.

Garrett, W.S., Lord, G.M., Punit, S., Lugo-Villarino, G., Mazmanian, S.K., Ito, S., Glickman, J.N., and Glimcher, L.H. (2007). Communicable ulcerative colitis induced by T-bet deficiency in the innate immune system. Cell 131, 33–45.

Kelly, C.P., and LaMont, J.T. (2008). Clostridium difficile—more difficult than ever. N. Engl. J. Med. 359, 1932–1940.

Garrett, W.S., Gallini, C.A., Yatsunenko, T., Michaud, M., DuBois, A., Delaney, M.L., Punit, S., Karlsson, M., Bry, L., Glickman, J.N., et al. (2010). Enterobacteriaceae act in concert with the gut microbiota to induce spontaneous and maternally transmitted colitis. Cell Host Microbe 8, 292–300.

Kinnebrew, M.A., Ubeda, C., Zenewicz, L.A., Smith, N., Flavell, R.A., and Pamer, E.G. (2010). Bacterial flagellin stimulates Toll-like receptor 5-dependent defense against vancomycin-resistant Enterococcus infection. J. Infect. Dis. 201, 534–543.

Glynn, J.R., Hornick, R.B., Levine, M.M., and Bradley, D.J. (1995). Infecting dose and severity of typhoid: analysis of volunteer data and examination of the influence of the definition of illness used. Epidemiol. Infect. 115, 23–30.

Kobayashi, K.S., Chamaillard, M., Ogura, Y., Henegariu, O., Inohara, N., Nun˜ez, G., and Flavell, R.A. (2005). Nod2-dependent regulation of innate and adaptive immunity in the intestinal tract. Science 307, 731–734.

Godinez, I., Haneda, T., Raffatellu, M., George, M.D., Paixa˜o, T.A., Rola´n, H.G., Santos, R.L., Dandekar, S., Tsolis, R.M., and Ba¨umler, A.J. (2008). T cells help to amplify inflammatory responses induced by Salmonella enterica serotype Typhimurium in the intestinal mucosa. Infect. Immun. 76, 2008–2017.

Kuwahara, T., Ogura, Y., Oshima, K., Kurokawa, K., Ooka, T., Hirakawa, H., Itoh, T., Nakayama-Imaohji, H., Ichimura, M., Itoh, K., et al. (2011). The lifestyle of the segmented filamentous bacterium: a non-culturable gut-associated immunostimulating microbe inferred by whole-genome sequencing. DNA Res. 18, 291–303.

Godinez, I., Raffatellu, M., Chu, H., Paixa˜o, T.A., Haneda, T., Santos, R.L., Bevins, C.L., Tsolis, R.M., and Ba¨umler, A.J. (2009). Interleukin-23 orchestrates mucosal responses to Salmonella enterica serotype Typhimurium in the intestine. Infect. Immun. 77, 387–398. Hadis, U., Wahl, B., Schulz, O., Hardtke-Wolenski, M., Schippers, A., Wagner, N., Mu¨ller, W., Sparwasser, T., Fo¨rster, R., and Pabst, O. (2011). Intestinal tolerance requires gut homing and expansion of FoxP3+ regulatory T cells in the lamina propria. Immunity 34, 237–246. Happel, K.I., Dubin, P.J., Zheng, M., Ghilardi, N., Lockhart, C., Quinton, L.J., Odden, A.R., Shellito, J.E., Bagby, G.J., Nelson, S., and Kolls, J.K. (2005). Divergent roles of IL-23 and IL-12 in host defense against Klebsiella pneumoniae. J. Exp. Med. 202, 761–769.

Lathrop, S.K., Bloom, S.M., Rao, S.M., Nutsch, K., Lio, C.W., Santacruz, N., Peterson, D.A., Stappenbeck, T.S., and Hsieh, C.S. (2011). Peripheral education of the immune system by colonic commensal microbiota. Nature, in press. Published online September 21, 2011. 10.1038/nature10434. Lawley, T.D., Clare, S., Walker, A.W., Goulding, D., Stabler, R.A., Croucher, N., Mastroeni, P., Scott, P., Raisen, C., Mottram, L., et al. (2009). Antibiotic treatment of clostridium difficile carrier mice triggers a supershedder state, sporemediated transmission, and severe disease in immunocompromised hosts. Infect. Immun. 77, 3661–3669. Lee, Y.K., Menezes, J.S., Umesaki, Y., and Mazmanian, S.K. (2011). Proinflammatory T-cell responses to gut microbiota promote experimental

Cell Host & Microbe 10, October 20, 2011 ª2011 Elsevier Inc. 321

Cell Host & Microbe

Review autoimmune encephalomyelitis. Proc. Natl. Acad. Sci. USA 108 (Suppl 1 ), 4615–4622. Lochner, M., Be´rard, M., Sawa, S., Hauer, S., Gaboriau-Routhiau, V., Fernandez, T.D., Snel, J., Bousso, P., Cerf-Bensussan, N., and Eberl, G. (2011). Restricted microbiota and absence of cognate TCR antigen leads to an unbalanced generation of Th17 cells. J. Immunol. 186, 1531–1537. Lupp, C., Robertson, M.L., Wickham, M.E., Sekirov, I., Champion, O.L., Gaynor, E.C., and Finlay, B.B. (2007). Host-mediated inflammation disrupts the intestinal microbiota and promotes the overgrowth of Enterobacteriaceae. Cell Host Microbe 2, 204. Macpherson, A.J., McCoy, K.D., Johansen, F.E., and Brandtzaeg, P. (2008). The immune geography of IgA induction and function. Mucosal Immunol. 1, 11–22. Maloy, K.J., and Powrie, F. (2011). Intestinal homeostasis and its breakdown in inflammatory bowel disease. Nature 474, 298–306.

Schulz, S.M., Ko¨hler, G., Holscher, C., Iwakura, Y., and Alber, G. (2008). IL-17A is produced by Th17, gammadelta T cells and other CD4- lymphocytes during infection with Salmonella enterica serovar Enteritidis and has a mild effect in bacterial clearance. Int. Immunol. 20, 1129–1138. Sczesnak, A., Segata, N., Qin, X., Gevers, D., Petrosino, J.F., Huttenhower, C., Littman, D.R., and Ivanov, I.I. (2011). The genome of th17 cell-inducing segmented filamentous bacteria reveals extensive auxotrophy and adaptations to the intestinal environment. Cell Host Microbe 10, 260–272. Sekirov, I., Tam, N.M., Jogova, M., Robertson, M.L., Li, Y., Lupp, C., and Finlay, B.B. (2008). Antibiotic-induced perturbations of the intestinal microbiota alter host susceptibility to enteric infection. Infect. Immun. 76, 4726– 4736. Sekirov, I., Gill, N., Jogova, M., Tam, N., Robertson, M., de Llanos, R., Li, Y., and Finlay, B.B. (2010). Salmonella SPI-1-mediated neutrophil recruitment during enteric colitis is associated with reduction and alteration in intestinal microbiota. Gut Microbes 1, 30–41.

Maslowski, K.M., Vieira, A.T., Ng, A., Kranich, J., Sierro, F., Yu, D., Schilter, H.C., Rolph, M.S., Mackay, F., Artis, D., et al. (2009). Regulation of inflammatory responses by gut microbiota and chemoattractant receptor GPR43. Nature 461, 1282–1286.

Simor, A.E., Bradley, S.F., Strausbaugh, L.J., Crossley, K., and Nicolle, L.E.; SHEA Long-Term-Care Committee. (2002). Clostridium difficile in long-termcare facilities for the elderly. Infect. Control Hosp. Epidemiol. 23, 696–703.

Mathis, D., and Benoist, C. (2011). Microbiota and autoimmune disease: the hosted self. Cell Host Microbe 10, this issue, 297–301.

Sonnenberg, G.F., Monticelli, L.A., Elloso, M.M., Fouser, L.A., and Artis, D. (2011). CD4(+) lymphoid tissue-inducer cells promote innate immunity in the gut. Immunity 34, 122–134.

Mazmanian, S.K., Round, J.L., and Kasper, D.L. (2008). A microbial symbiosis factor prevents intestinal inflammatory disease. Nature 453, 620–625. Niess, J.H., Brand, S., Gu, X., Landsman, L., Jung, S., McCormick, B.A., Vyas, J.M., Boes, M., Ploegh, H.L., Fox, J.G., et al. (2005). CX3CR1-mediated dendritic cell access to the intestinal lumen and bacterial clearance. Science 307, 254–258. O’Connor, J.R., Johnson, S., and Gerding, D.N. (2009). Clostridium difficile infection caused by the epidemic BI/NAP1/027 strain. Gastroenterology 136, 1913–1924. Petnicki-Ocwieja, T., Hrncir, T., Liu, Y.J., Biswas, A., Hudcovic, T., TlaskalovaHogenova, H., and Kobayashi, K.S. (2009). Nod2 is required for the regulation of commensal microbiota in the intestine. Proc. Natl. Acad. Sci. USA 106, 15813–15818. Prakash, T., Oshima, K., Morita, H., Fukuda, S., Imaoka, A., Kumar, N., Sharma, V.K., Kim, S.W., Takahashi, M., Saitou, N., et al. (2011). Complete genome sequences of rat and mouse segmented filamentous bacteria, a potent inducer of th17 cell differentiation. Cell Host Microbe 10, 273–284. Raffatellu, M., George, M.D., Akiyama, Y., Hornsby, M.J., Nuccio, S.P., Paixao, T.A., Butler, B.P., Chu, H., Santos, R.L., Berger, T., et al. (2009). Lipocalin-2 resistance confers an advantage to Salmonella enterica serotype Typhimurium for growth and survival in the inflamed intestine. Cell Host Microbe 5, 476–486. Rakoff-Nahoum, S., Paglino, J., Eslami-Varzaneh, F., Edberg, S., and Medzhitov, R. (2004). Recognition of commensal microflora by toll-like receptors is required for intestinal homeostasis. Cell 118, 229–241. Reid, D.M., Gow, N.A., and Brown, G.D. (2009). Pattern recognition: recent insights from Dectin-1. Curr. Opin. Immunol. 21, 30–37. Round, J.L., and Mazmanian, S.K. (2009). The gut microbiota shapes intestinal immune responses during health and disease. Nat. Rev. Immunol. 9, 313–323. Round, J.L., Lee, S.M., Li, J., Tran, G., Jabri, B., Chatila, T.A., and Mazmanian, S.K. (2011). The Toll-like receptor 2 pathway establishes colonization by a commensal of the human microbiota. Science 332, 974–977. Rupnik, M., Wilcox, M.H., and Gerding, D.N. (2009). Clostridium difficile infection: new developments in epidemiology and pathogenesis. Nat. Rev. Microbiol. 7, 526–536.

Spits, H., and Di Santo, J.P. (2011). The expanding family of innate lymphoid cells: regulators and effectors of immunity and tissue remodeling. Nat. Immunol. 12, 21–27. Stecher, B., Robbiani, R., Walker, A.W., Westendorf, A.M., Barthel, M., Kremer, M., Chaffron, S., Macpherson, A.J., Buer, J., Parkhill, J., et al. (2007). Salmonella enterica serovar typhimurium exploits inflammation to compete with the intestinal microbiota. PLoS Biol. 5, 2177–2189. Stecher, B., Chaffron, S., Ka¨ppeli, R., Hapfelmeier, S., Freedrich, S., Weber, T.C., Kirundi, J., Suar, M., McCoy, K.D., von Mering, C., et al. (2010). Like will to like: abundances of closely related species can predict susceptibility to intestinal colonization by pathogenic and commensal bacteria. PLoS Pathog. 6, e1000711. Suau, A., Bonnet, R., Sutren, M., Godon, J.J., Gibson, G.R., Collins, M.D., and Dore´, J. (1999). Direct analysis of genes encoding 16S rRNA from complex communities reveals many novel molecular species within the human gut. Appl. Environ. Microbiol. 65, 4799–4807. Sutton, C.E., Lalor, S.J., Sweeney, C.M., Brereton, C.F., Lavelle, E.C., and Mills, K.H. (2009). Interleukin-1 and IL-23 induce innate IL-17 production from gammadelta T cells, amplifying Th17 responses and autoimmunity. Immunity 31, 331–341. Tsuji, M., Suzuki, K., Kitamura, H., Maruya, M., Kinoshita, K., Ivanov, I.I., Itoh, K., Littman, D.R., and Fagarasan, S. (2008). Requirement for lymphoid tissueinducer cells in isolated follicle formation and T cell-independent immunoglobulin A generation in the gut. Immunity 29, 261–271. Ubeda, C., Taur, Y., Jenq, R.R., Equinda, M.J., Son, T., Samstein, M., Viale, A., Socci, N.D., van den Brink, M.R., Kamboj, M., and Pamer, E.G. (2010). Vancomycin-resistant Enterococcus domination of intestinal microbiota is enabled by antibiotic treatment in mice and precedes bloodstream invasion in humans. J. Clin. Invest. 120, 4332–4341. Umesaki, Y., Okada, Y., Matsumoto, S., Imaoka, A., and Setoyama, H. (1995). Segmented filamentous bacteria are indigenous intestinal bacteria that activate intraepithelial lymphocytes and induce MHC class II molecules and fucosyl asialo GM1 glycolipids on the small intestinal epithelial cells in the ex-germ-free mouse. Microbiol. Immunol. 39, 555–562. Vaishnava, S., Behrendt, C.L., Ismail, A.S., Eckmann, L., and Hooper, L.V. (2008). Paneth cells directly sense gut commensals and maintain homeostasis at the intestinal host-microbial interface. Proc. Natl. Acad. Sci. USA 105, 20858–20863.

Sadlack, B., Merz, H., Schorle, H., Schimpl, A., Feller, A.C., and Horak, I. (1993). Ulcerative colitis-like disease in mice with a disrupted interleukin-2 gene. Cell 75, 253–261.

van de Pavert, S.A., and Mebius, R.E. (2010). New insights into the development of lymphoid tissues. Nat. Rev. Immunol. 10, 664–674.

Schneider, D.S., and Ayres, J.S. (2008). Two ways to survive infection: what resistance and tolerance can teach us about treating infectious diseases. Nat. Rev. Immunol. 8, 889–895.

van der Waaij, D., Berghuis-de Vries, J.M., and Lekkerkerk Lekkerkerk-v. (1971). Colonization resistance of the digestive tract in conventional and antibiotic-treated mice. J. Hyg. (Lond.) 69, 405–411.

322 Cell Host & Microbe 10, October 20, 2011 ª2011 Elsevier Inc.

Cell Host & Microbe

Review Vollaard, E.J., and Clasener, H.A. (1994). Colonization resistance. Antimicrob. Agents Chemother. 38, 409–414. Winter, S.E., Thiennimitr, P., Winter, M.G., Butler, B.P., Huseby, D.L., Crawford, R.W., Russell, J.M., Bevins, C.L., Adams, L.G., Tsolis, R.M., et al. (2010). Gut inflammation provides a respiratory electron acceptor for Salmonella. Nature 467, 426–429. Wlodarska, M., Willing, B., Keeney, K.M., Menendez, A., Bergstrom, K.S., Gill, N., Russell, S.L., Vallance, B.A., and Finlay, B.B. (2011). Antibiotic treatment alters the colonic mucus layer and predisposes the host to exacerbated Citrobacter rodentium-induced colitis. Infect. Immun. 79, 1536–1545.

Wu, H.J., Ivanov, I.I., Darce, J., Hattori, K., Shima, T., Umesaki, Y., Littman, D.R., Benoist, C., and Mathis, D. (2010). Gut-residing segmented filamentous bacteria drive autoimmune arthritis via T helper 17 cells. Immunity 32, 815–827.

Zheng, Y., Valdez, P.A., Danilenko, D.M., Hu, Y., Sa, S.M., Gong, Q., Abbas, A.R., Modrusan, Z., Ghilardi, N., de Sauvage, F.J., and Ouyang, W. (2008). Interleukin-22 mediates early host defense against attaching and effacing bacterial pathogens. Nat. Med. 14, 282–289.

Cell Host & Microbe 10, October 20, 2011 ª2011 Elsevier Inc. 323