Pathogen Colonization Resistance in the Gut and Its Manipulation for Improved Health

Abstract

Mammals have coevolved with a large community of symbiotic, commensal, and some potentially pathogenic microbes. The trillions of bacteria and hundreds of species in our guts form a relatively stable community that resists invasion by outsiders, including pathogens. This powerful protective force is referred to as colonization resistance. We discuss the variety of proposed or demonstrated mechanisms that can mediate colonization resistance and some potential ways to manipulate them for improved human health. Instances in which certain bacterial pathogens can overcome colonization resistance are also discussed.

The bacteria and other microbes inhabiting the animal gut (the gut microbiota) provide major beneficial functions to the host. These include nutrition (breakdown of indigestible polysaccharides and production of vitamins) and protection against microbes that could harm the host. The idea that certain good bacteria could compete with harmful ones was first proposed by Élie Metchnikoff in the early 20th Century. With the advent of antibiotics and germ-free animals, it was demonstrated that the resident bacteria in the gut play a large role in preventing pathogens from colonizing and causing disease, in animals and humans.¹ This phenomenon was later coined colonization resistance.² In addition to defense against strict pathogens, the same concept applies to control of indigenous but potentially dangerous pathobionts, as well as exclusion of innocuous foreign species, such as probiotics. There are a wide variety of mechanisms now known to participate in colonization resistance. Many involve direct interactions between bacterial cells, whereas others act by modulating host physiology, especially the immune system. In addition, host genetics, diet, and antibiotic use can modify the composition and function of the microbiota and, thus, affect colonization resistance. In the clinic, fecal microbiota transplantation is used effectively to treat antibiotic-induced diarrhea caused by the overgrowth of Clostridium difficile.³ We will discuss the history and new developments in understanding this phenomenon as it relates to enteric pathogens, probiotics, prebiotics, and improving human health.

Direct Mechanisms

Bacteria can reach densities approaching 10¹¹ cells/g in the mammalian large intestine⁴ and are in constant competition with each other for survival. Nutrients, including carbon, nitrogen, and energy sources, as well as other essential molecules can be limited, especially in the large intestine. As a consequence of this highly competitive environment, bacteria have evolved ways to suppress or kill each other. These direct mechanisms of resistance are summarized in Figure 1.

Figure 1.

Direct mechanisms. Anaerobic symbionts (top left) digest host mucus and dietary polysaccharides, releasing monosaccharides for another symbiont to take up, excreting short-chain fatty acids (SCFAs), and acidifying the environment, which suppresses growth of a pathogen. Bacteria deconjugate bile acids (top right), and others produce 7α-dehydroxylases that convert them to secondary bile acids, which also inhibit the pathogen. A symbiont produces bacteriocins (green), which form pores in the pathogen, allowing leakage of cellular contents (gray). Symbionts could also theoretically target pathogens with contact-dependent mechanisms, like the type 6 secretion system (T6SS; blue) or contact-dependent inhibition (CDI; red).

Nutritional Competition

Monosaccharides, disaccharides, and most proteins ingested by mice or humans are digested and absorbed in the small intestine. This leaves only complex plant polysaccharides and host-secreted mucus as food sources for the dense community of bacteria in the large intestine. Metabolizing these polysaccharides requires specific enzymes. Some bacteria that predominate in the large intestine, such as Bacteroides species, possess a vast array of genes to harvest sugars from host and diet polysaccharides.⁵ Others, like the often pathogenic Enterobacteriaceae, can generally only use simpler sugars and amino acids for their carbon, nitrogen, and energy needs; and they are found at low levels in the healthy gut. When resident bacteria are acutely killed with an antibiotic, there is a surplus of free monosaccharides released from host glycans, such as sialic acid and fucose, which are taken advantage of by some pathogens, like Salmonella enterica serovar Typhimurium (S. typhimurium).⁶ Increased sugars and free amino acids available after antibiotic treatment may also be exploited by C. difficile, a major human pathogen that causes pseudomembranous colitis.6, 7, 8, 9 Constant scavenging of all available nutrients in the normal gut may, therefore, prevent invasion and maintain a stable beneficial community. On the basis of in vitro experiments, Freter et al proposed that the steady-state abundance of different taxa in the gut was determined by “one or a few nutritional substrates which a given strain can utilize most efficiently.”^10,pp.676 This nutrient niche hypothesis is supported by in vitro and in vivo experiments.11, 12, 13 When a single host-produced sugar, fucose, is removed from the mouse gut, the microbiota does change in composition,¹⁴ but beyond that, the microbiota's ability to suppress pathogens and pathobionts is altered.15, 16, 17 On the other hand, removal of a host-supplied sugar will also make it unavailable to pathogens or pathobionts, even when nutrient scavenging is disrupted, and therefore can improve resistance (as in the case of sialic acid).¹⁸ Interestingly, transgenic expression in mice of a sugar structure not normally found in their stomach was also able to affect the phenotype of the pathobiont Helicobacter pylori.¹⁹ The host sugar structures (glycosylation) in the gut can be modified by a variety of signals, including microbial colonization,²⁰ immune cell activation,15, 17 and various chemicals,²¹ so manipulation of the host-produced nutrients in the gut could be a way to alter the microbiota composition or function.

Diet can have a rapid and profound impact on microbiota composition and function and, consequently, colonization resistance.²² Complex plant polysaccharides (fiber) in the diet are a major food source for the anaerobic bacteria that dominate the lower gastrointestinal tract. Removing fermentable polysaccharides from the diet causes a shift in overall community structure and, over time, can lead to permanent loss of species from the gut.²³ Loss of protective species or hampering of their normal functions in the absence of their preferred food sources could lead to reduced colonization resistance. For example, when deprived of diet polysaccharides, bacteria switch feeding preferences toward mucus glycans and proteins, degrading them, reducing the effectiveness of this protective barrier, and increasing the damage caused by Citrobacter rodentium.²⁴ Feeding a polysaccharide-free diet also allowed easier invasion of C. difficile, which could be rescued by supplementation with inulin, a fructose-based polysaccharide found in many plants.²⁵ Rice bran added to an otherwise low-fiber diet reduced S. typhimurium colonization.²⁶ However, another study of several purified polysaccharides found that most did not reduce S. typhimurium loads and, in fact, some had the opposite effect.²⁷ It is important to take into account the exact source and structure of polysaccharides (of which there are an almost infinite variety), the dietary context, and the preexisting microbiota and its capacity to respond to a given substrate. Other plant-derived compounds in the diet, besides polysaccharides, could also have myriad effects on both direct and indirect colonization resistance.²⁸ Western diets, which are higher in fat and simple sugars but low in complex polysaccharides, favor the expansion of endogenous Proteobacteria, such as Escherichia coli, and allow easier introduction of a human pathobiont associated with inflammatory bowel disease, adherent-invasive E. coli.²⁹ As discussed later, high-fat diets can also alter the microbiota and colonization resistance via bile acid production, as well as having indirect effects on the immune system. Low-protein diets, on the contrary, can enhance protection against C. difficile, possibly by reducing the free amino acids available to it.7, 30 Of course, malnutrition will have many detrimental effects on host functions, but permanent changes to the microbiota are also apparent in malnourished children.³¹ This could mean that there are additional microbiota-intrinsic defects in colonization resistance in these already vulnerable individuals. Because diet can have innumerable effects on host physiology, separating the indirect effects from direct influences on the microbiota is not a simple task.

Prebiotics are dietary nutrients, typically polysaccharides, used to target subsets of the indigenous microbiota and bolster their beneficial functions. Identifying prebiotics, such as inulin,²⁵ that specifically enhance colonization resistance and the mechanisms behind this could have immense clinical benefits. Nutrient niches can even be artificially generated in the gut via diet. Two recent studies demonstrated that, by introducing a specific polysaccharide in the diet (porphyran from seaweed) along with a bacterial strain uniquely capable of using it, the strain could be stably inserted into the gut community.32, 33

If nutritional competition is responsible for steady-state microbiota structure and exclusion of outsiders, are all functional niches constantly filled in a healthy gut? This may not always be the case in humans. In one study, subjects were fed a probiotic, Bifidobacterium longum, and then examined for persistence of the strain in their stool. Although probiotics generally do not stably colonize in humans, surprisingly approximately a third of the subjects maintained detectable levels of the strain up to 200 days later.³⁴ Before treatment, the persisters generally had lower amounts of endogenous B. longum in their gut and/or lower amounts of B. longum genes, including some related to carbohydrate use. This suggests that, for unknown reasons, some individuals had an open functional niche that the strain was able to fill. Another recent study gave subjects a mix of 11 probiotics and also found individuals who had persistence of some strains in the large intestinal mucus layer.³⁵ Persistence seemed to be controlled by differences in their microbiota because the result could be recapitulated by transfer to mice, but the underlying mechanisms that control persistence of probiotics are not yet clear.

Overall, nutritional competition seems to be a powerful force for excluding nonnative bacteria from the gut ecosystem, including beneficial probiotics. Fortunately, the nutritional landscape in the gut can be easily modified via diet. Prebiotics, such as plant polysaccharides, may be used to target specific endogenous beneficial species and enhance their functions. Probiotic strains, delivered along with prebiotics that they can use (a combination known as a synbiotic), could facilitate introduction of beneficial strains to the existing gut community.

Bactericidal/Bacteriostatic Mechanisms

Perhaps because competition for growth substrates is so intense, gut bacteria have also developed many ways of suppressing or killing competitors. Indeed, dominant growth suppression may be more important than nutrient limitation for colonization resistance in some cases.³⁶ Bacteriocins, for instance, are a large, heterogeneous group of peptides produced by bacteria, with diverse bacteriostatic or bactericidal activities.³⁷ Bacteriocin production is common in bacteria used as probiotics, fermented foods, and the gut microbiota, and many have been developed as possible replacements for traditional antibiotics. How important is production of antibacterial molecules for competition in the gut? Production or susceptibility to colicins (a family of bacteriocins) affected E. coli competition in antibiotic-treated mouse models,38, 39 but there is also evidence that they mediate ongoing competition among E. coli strains in the natural microbiota as well.⁴⁰ A bacteriocin of the common gut resident Enterococcus faecalis gave it a competitive advantage against other Enterococci in mice and allowed it to repel a related pathogenic species, vancomycin-resistant Enterococcus faecium.⁴¹ Bacteroides species, which are highly abundant in mouse and human gut, produce many bactericidal or bacteriostatic secreted proteins as well, some of which have been shown to mediate intraspecies competition in vivo.⁴²

Because of their relative ease of discovery and therapeutic utility, most of the antibacterial factors known in the gut are secreted, soluble compounds. However, mechanisms of suppression that require direct cell-cell contact are increasingly being appreciated as well. The contact-dependent inhibition system was discovered in E. coli,⁴³ and homologous genes are distributed throughout the Proteobacteria and possibly other phyla as well.⁴⁴ The contact-dependent inhibition consists of two genes that compose a two-partner secretion system (in the type 5 secretion system family). They assemble a long filament that extends out from the cell. This binds target cells via a specific receptor and delivers the inhibitory activity found in the filament protein's effector domain. This effector domain is variable between strains and interchangeable, and effectors have been found with DNase, RNase, and pore-forming activities. A cognate immunity protein is encoded along with the system, to inactivate the effector and prevent self-inhibition.

The type 6 secretion system (T6SS) is another recently recognized player in contact-dependent competition in the gut. The T6SS was originally identified in several Gram-negative species as a secretion system that was involved in interactions with eukaryotic cells.45, 46, 47, 48 Subsequently, this was extended to interbacterial action as well⁴⁹ and, specifically, intraspecies killing.⁵⁰ The T6SS system works by spearing nearby cells and delivering effectors into their cytoplasm. These effectors can degrade nucleotides, cell walls, or membranes, or they have other activities. Like the contact-dependent inhibition, the effectors are often found with a cognate immunity protein. The T6SS genes are widespread in Gram-negative bacteria, especially Proteobacteria, and a related but distinct family exists in the Bacteroidales.51, 52 More important, T6SS effector and immunity genes may contribute to ongoing competition among the abundant Bacteroides species in the mouse and human gut.53, 54, 55 Some pathogens can also use their T6SS to kill resident bacteria and enable their invasion of the gut community.56, 57 Another secretion system, called Esx or type 7, can mediate intraspecies and interspecies killing by Gram-positive bacteria.58, 59 Notably, at least one toxin family secreted by this system is abundant among the Firmicutes, such as Clostridia and Bacilli, in the human microbiome.⁵⁹

Contact-dependent systems of growth inhibition and killing are increasingly being discovered in the gut microbiota and could likely participate in resistance to pathogens. The modularity of effector/immunity genes common to these systems may make them amenable to engineering. These contact-dependent mechanisms also highlight the value of studying the microscopic structure and spatial relationships in the gut community.⁶⁰ However, the overall role of bactericidal/bacteriostatic mechanisms in mediating colonization resistance against pathogens, pathobionts, and probiotics remains poorly understood.

Metabolites and Chemical Transformations

Bacteria also produce metabolic by-products or modify compounds in the gut that can affect their growth. Short-chain fatty acids (SCFAs) are a major product of bacterial fermentation in the large intestine, especially downstream of polysaccharide digestion, and they can have growth-inhibiting effects on pathogenic bacteria, such as E. coli, S. typhimurium, and C. difficile.61, 62, 63 Caused, in part, by SCFA excretion, microbial metabolism generally lowers pH in the gut, especially in the large intestine.⁶⁴ Oxygen is normally scarce in the large intestine lumen, which has been attributed to bacterial and host metabolism, the latter driven by epithelial catabolism of the SCFA butyrate.65, 66 Dysbiosis, a general deviation in community structure associated with disease and breakdown of colonization resistance, frequently coincides with increased oxygen availability, expansion of facultative anaerobes, such as pathogenic Proteobacteria,⁶⁷ and reduction in butyrate and butyrate producers. Thus, prebiotic polysaccharides that enhance SCFA production could improve colonization resistance in several possible ways.

Bile acids are produced in the liver from cholesterol, stored in the gallbladder, and secreted into the duodenum after eating. Their amphipathic qualities help dissolve fat and fat-soluble vitamins for absorption, but they are antibacterial to varying degrees as well. The primary bile acids in humans are cholic acid and chenodeoxycholic acid, which are synthesized in the liver and conjugated to taurine or glycine. Once in the gut, bile salt hydrolases, produced by many different bacterial taxa, deconjugate them from the amino acid, possibly to reduce their toxicity or to obtain the taurine or glycine itself. Once deconjugated, the primary bile acids can be converted into a variety of secondary bile acids by enzymes produced by rare bacterial species in the gut.⁶⁸ 7α-Dehydroxylation, for example, can convert cholic acid to deoxycholic acid and chenodeoxycholic acid to lithocholic acid. Both of these secondary bile acids can suppress growth of C. difficile.⁶⁹ At the same time, the conjugated primary bile acid taurocholic acid promotes C. difficile spore germination.⁷⁰ Thus, microbial transformation of bile acids may partially explain susceptibility to C. difficile.9, 69, 71 Bile acids are also modulated by diet: excessive taurocholic acid, induced by a high-fat diet, promoted expansion of a pathobiont, Bilophila wadsworthia, leading to worsened colitis in that model.⁷² Although impacts on fat absorption and metabolism must be considered, modulating the levels of bile acids via diet, specific bacteria,69, 73 or drugs could be a means to enhance colonization resistance.

Indirect Mechanisms and Modulation of Host Physiology

Although the fierce ongoing competition between bacteria in the gut is largely responsible for maintaining a healthy, beneficial, and stable community and effectively repelling invaders, the host organism also has a role to play. The host can shape the structure and activity of the microbiota, thereby affecting its functions, including its level of colonization resistance. At the same time, the host immune system and other aspects of its biology are heavily influenced by the microbiota (Figure 2).

Figure 2.

Indirect mechanisms. Dimeric IgA, produced by a B cell in the lamina propria, is transcytosed by the poly-Ig receptor into the lumen, where it binds a bacterium's flagella. Farnesoid X receptor (FXR) and TGR5 on epithelial cells and macrophages up-regulate defenses or modulate inflammation, respectively, in response to bile acids. The short-chain fatty acid butyrate, produced by anaerobic Clostridia, promotes oxygen respiration in an epithelial cell, reducing the oxygen concentration at the epithelial surface. Toll-like receptors (TLRs) on epithelial cells, macrophages, and dendritic cells (DCs) can sense microbial molecules and signal through myeloid differentiation primary response 88 (MyD88) or TIR-domain-containing adapter-inducing interferon-β (TRIF) adaptors. Macrophages make IL-1β and DCs make IL-23 cytokines in response to TLR stimulation, which induces type 17 helper T cells (Th17s) and innate-like lymphocytes (ILCs) to secrete IL-22. This acts on epithelial cells, causing them to produce regenerating islet-derived protein 3β (Reg3β) (red) and lipocalin-2 (green), which attack a pathogen in the lumen and sequester iron (red circles) from it, respectively. On the right side, protective symbionts, like Clostridia, have been depleted (eg, by antibiotics), resulting in increased oxygen in the lumen. Salmonella enterica serovar Typhimurium (S. typhimurium) is resistant to Reg3β and can capture iron from lipocalin-2. Meanwhile Citrobacter rodentium uses its type 3 secretion system to inject effectors into an epithelial cell and cause hyperplasia, further increasing oxygen levels and supporting its replication.

Innate Defense Mechanisms

Factors that can prevent or promote growth of microbes and, thus, select which bacteria can colonize mucosal surfaces, like the gut, are evidently ancient. The hydra, for example, an extremely simple type of aquatic animal, nonetheless selects and maintains a species-specific group of bacterial symbionts. This is accomplished, in part, through production of different antibacterial compounds.⁷⁴ Transferring microbiota between more complex animals (eg, zebra fish and mice) has confirmed that the host significantly shapes the bacterial community in the gut.⁷⁵ Many mammalian genes are regulated by microbial colonization, including antibacterial factors.⁷⁶ Some of these genes are regulated through sensing pathways of the innate immune system: toll-like receptors (TLRs), which signal through the adaptor myeloid differentiation primary response 88 (MyD88) or TIR-domain-containing adapter–inducing interferon-β (TRIF), or nucleotide-binding oligomerization domain–containing protein (NOD)-like receptors, which sense microbial molecules in the cytosol and assemble inflammasomes.⁷⁷ These receptors sense conserved microbial molecules, such as cell wall or outer membrane components, which are present in both symbionts and pathogens. TLRs and NOD-like receptors, as well as other microbial pattern sensors, can be expressed by epithelial cells and hematopoietic cells, such as macrophages and dendritic cells. Steady-state sensing of the microbiota through TLRs does occur and regulates many genes in the gut as the microbiota develops.⁷⁸ Although many of these are antimicrobial genes, the impact of TLR signaling on the steady-state composition of the microbiota seems to be minimal in well-controlled experiments.⁷⁸ However, the system can be used to artificially boost resistance. Systemic injection of TLR ligands, for instance, activates a signaling pathway that culminates in production of antimicrobial peptides and protects against vancomycin-resistant E. faecium.⁷⁹ Dendritic cells in the intestinal lamina propria are activated through their TLRs to produce IL-23. This causes production of another cytokine, IL-22, which is crucial for epithelial defense and repair in the gut,⁸⁰ and can be produced by innate-like lymphocytes or type 17 helper T cells. The IL-22 induces antimicrobial peptides, particularly regenerating islet-derived protein 3β (Reg3β) and Reg3γ, in epithelial cells, resulting in reduced lumen colonization and invasion of vancomycin-resistant E. faecium.⁷⁹ However, no protective role for MyD88/TRIF signaling or IL-22 was observed in colonization resistance against S. typhimurium.⁸¹ In fact, effectors induced by IL-22 may give S. typhimurium an advantage over other bacteria in the gut.⁸² Further studies are needed to understand the physiological role of IL-22 and antimicrobial peptides in the regulation of pathogen colonization resistance.

Bacterial metabolites can also have a multitude of effects on host physiology and, especially, immune system function through many different signaling pathways.⁸³ SCFAs, products of polysaccharide digestion, can affect both innate inflammation as well as B- and T-cell differentiation. Bile acids that have been modified by bacteria can activate antimicrobial genes in the ileum via the farnesoid X receptor or modulate inflammation in immune cells through another receptor, TGR5.⁸³ The aryl hydrocarbon receptor senses microbial and diet-derived molecules and can regulate epithelial, dendritic cell, T-cell, and innate-like lymphocyte function.⁸³ These are only a few known examples of the large variety of microbiota-derived metabolites that can potentially affect the host animal.

B Cells, Igs, and Adaptive Immunity

The adaptive immune system may add another layer of specificity to control of the microbiota. Humans secrete several grams of IgA daily into the gut. Most of the bacterial cells in the small intestine are coated with IgA, with less coating in the large intestine and feces.84, 85 IgA production is driven by the microbiota, being lower overall in germ-free mice but induced by introduction of certain bacteria or bacterial products.86, 87 Most of the secreted IgA in the steady state is produced by B cells independently of T-cell help and is relatively low affinity but polyreactive, in that a single antibody can bind to more than one bacterial species or even phylum.84, 88, 89, 90 This microbiota-stimulated, secreted IgA could theoretically mediate colonization resistance in two ways: by direct binding to a pathogen/pathobiont or by modulating the composition or function of the resident microbiota.

Indeed, IgA may preferentially target dangerous bacteria in the normal gut community (ie, pathobionts) because transferring the IgA-coated population from humans with inflammatory bowel disease to mice worsened disease in a colitis model.⁹¹ IgA also targets flagellar proteins in the normal microbiota, down-regulating their expression and reducing motility/invasion.⁹² These antibodies can cross-react with pathogens, such as S. typhimurium, and so serve as a preformed line of defense. Indeed, the amount of preexisting IgA that bound S. typhimurium differed between two common inbred mouse strains, and higher amounts correlated with better survival after oral infection, suggesting that this preexisting pathogen-reactive antibody could support colonization resistance.⁹³

Although IgA is actively transported into the gut lumen, IgG antibodies do not normally cross the barrier unless it is damaged. However, there is IgG in the serum that reacts against gut bacteria.94, 95 This antibody may be stimulated by and help to control the normal low-level translocation of bacteria from the gut. If the epithelial barrier is breached by a pathogen or damaged, the premade IgG is ready to help control infection.⁹⁴

How large a role secreted IgA plays in shaping the normal gut microbiota remains controversial. Mice that lack secreted IgA for various reasons have been reported to have widely varying amounts of dysbiosis.96, 97 Similarly, mice that lack the polymeric Ig receptor and cannot transport IgA or IgM into the gut lumen were found to be either more or less susceptible to S. typhimurium infection than wild-type mice.98, 99 The reason for these conflicting results is unclear, but they are likely explained by underlying microbiota differences.

Aside from changing community composition, could IgA that coats symbiotic bacteria modulate their protective function? IgA against the symbiont Bacteroides thetaiotaomicron down-regulated its expression of the target epitope and reduced its fitness in the gut.¹⁰⁰ Even binding of a nonspecific IgA, likely via glycosylation on the constant region, can modulate B. thetaiotaomicron gene expression in the gut.¹⁰¹ Although antibody binding is usually thought of as having an exclusionary function, IgA surprisingly promotes the survival of some symbionts in humans and mice, perhaps by retaining them in a favorable physical niche.102, 103

Selective IgA deficiency in humans is relatively common (approximately 0.1% to 1%), and although it often goes unnoticed, it does increase the overall risk of infection (especially respiratory and gastrointestinal tract) and autoimmunity, including celiac disease and inflammatory bowel disease,104, 105 which are tentatively linked to pathogen or pathobiont infections, respectively. In the absence of IgA, IgM can also be secreted in comparable quantities and bind to similar bacterial populations but may not compensate fully.84, 103

Although mostly dispensable for steady-state IgA, T cells can affect colonization resistance via cytokine production. Type 17 helper T cells, for instance, are produced in response to some resident bacteria, including segmented filamentous bacteria, a unique species that attaches directly to intestinal epithelial cells.¹⁰⁶ IL-1β production by macrophages, induced by the microbiota via TLRs, stimulates type 17 helper T-cell development.¹⁰⁷ The type 17 helper T cells produce IL-22, which induces antibacterial and tissue repair genes in the epithelium and results in improved resistance to C. rodentium.¹⁰⁶

Indirect mechanisms of colonization resistance, which are induced in the host by the microbiota, are exceedingly difficult to dissect away from the influence of the bacteria directly. Although it is clear that the host organism has some ability to shape its microbiota, through primitive innate mechanisms as well as more specific adaptive ones, it seems that we are largely bystanders when it comes to colonization resistance.

Pathogen Evasion of Colonization Resistance

Colonization resistance is generally effective: for example, large doses (≥10¹¹ is not uncommon) of different probiotics are regularly ingested by humans and usually do not establish in the gut. The dose of S. typhimurium needed to successfully infect mice decreases by many orders of magnitude after treatment with an antibiotic.¹ Clostridium difficile colitis in humans is also strongly associated with prior antibiotic use.¹⁰⁸ Despite this, some pathogens seem to have developed ways to partially circumvent colonization resistance. One is C. rodentium, a mouse-specific pathogen that infects the large intestine, causes colonic hyperplasia, and is then cleared within 3 weeks (in most mouse strains). The dose required for a successful infection with C. rodentium is fairly low when spreading naturally between mice.¹⁰⁹ Most important, once an infection is established, the levels of C. rodentium shed into the large intestine lumen and feces quickly increase to 10⁹/g, much higher than any endogenous Proteobacteria. How does it achieve this?

Citrobacter rodentium is similar to the human diarrhea-causing pathogens enteropathogenic and enterohemorrhagic E. coli. All three pathogens exhibit a similar specialized lifestyle in the gut: direct attachment to the epithelium and formation of lesions (giving them the name attaching/effacing pathogens). This attaching/effacing behavior is made possible by a cluster of virulence genes that the strains share, called the locus of enterocyte effacement. The locus of enterocyte effacement comprises a type 3 secretion system and associated effectors, as well as an essential surface protein, intimin. The ability of C. rodentium to live and replicate at the epithelial surface likely lets it avoid competition with most bacteria in the large intestine lumen and, therefore, the main force of colonization resistance.

The epithelial cell surface is already higher in oxygen than the lumen, which excludes the obligate anaerobes that make up most of the microbiota. Citrobacter rodentium further increases oxygenation by triggering epithelial hyperplasia via type 3 secretion system–injected effectors.¹¹⁰ Although it does not possess enzymes to digest mucus glycans, there may be other unique nutrient sources in this epithelial niche that C. rodentium exploits. Once the host initiates an IgG antibody response against locus of enterocyte effacement virulence factors, including surface intimin, opsonized virulent C. rodentium are removed by luminal neutrophils while avirulent bacteria are forced away from the protected niche at the epithelial surface and into the lumen. Without the advantage of their virulence factors, they are unable to compete with the microbiota and are cleared from the gut.¹¹¹ This clearance is probably caused, in part, by nutrient competition with other bacteria in the lumen.¹¹¹

Salmonella typhimurium exhibits a different lifestyle in mice than attaching/effacing pathogens. Unlike the diarrheal infection it causes in humans, S. typhimurium in mice is largely focused on invasion across the epithelium and into systemic organs, causing a typhoid-like disease without intestinal inflammation. However, if colonization resistance is ablated with antibiotics, it deploys several strategies to compete with the microbiota in the gut. Like C. rodentium, it uses type 3 secretion systems to target host epithelial and immune cells and induces inflammation and oxygenation of the cecum. This results in new energy and nutrient sources that it can exploit better than resident microbes.66, 112, 113 It has also evolved resistance to some host antimicrobial factors (eg, Reg3β¹¹⁴ and lipocalin-2).¹¹⁵ Similar to C. rodentium, its eventual clearance from the gut requires a fully functional microbiota.¹¹⁶ Whether humans exhibit a different disease from S. typhimurium because of differences in basic physiology or the microbiota is still unclear, although interestingly the microbiota from a healthy human did not protect mice as effectively as their native bacteria.¹¹⁷

Conclusions and Future Perspectives

Colonization resistance is a powerful phenomenon that arises from the extremely complex interactions of thousands of strains of bacteria in the gut with each other and with the host animal. Understanding the mechanisms of colonization resistance (Figures 1 and 2) is crucial to preventing and treating human disease. Approximately half a million children die each year from diarrheal diseases, for example. The infant microbiota is not fully developed and is intrinsically defective in colonization resistance ability.⁸¹ Maturation of the microbiota to a protective adult state in mice occurs around the time that solid food is starting to be introduced. This suggests that dietary interventions could be a way to boost the development of colonization resistance in infants. Synbiotic combinations of protective bacteria and prebiotics that support them could be even more powerful, if safe and effective strains can be isolated. Infants also have immature immune systems, and so they are defective in indirect mechanisms of colonization resistance, such as antibody production. This gap is filled by maternal antibodies transferred across the placenta and in breast milk but could be enhanced by maternal immunization.

The arsenal of weapons that bacteria use to compete with each other has historically been and continues to be a source of useful compounds. Discovering new types of antibacterials is especially important now as resistance to common antibiotics continues to increase, and the adverse effects of broad-spectrum antibiotics on the microbiota are increasingly appreciated. Discovering antibacterials from the gut microbiome that are more specific and with fewer adverse effects is another advantage of understanding colonization resistance.

Transfers of the total microbiota (fecal microbiota transplantation) are more effective than antibiotics at treating C. difficile infection,³ and they are being explored for other diseases as well. This presents an interesting opportunity to try to predict and understand why certain species from the donor can successfully colonize the recipient.¹¹⁸ Fecal microbiota transplantations could reveal mechanisms controlling normal microbiota stability and exclusion of nonnative strains.

Colonization resistance is not only a fascinating and complex phenomenon that incorporates aspects of ecology, microbiology, biochemistry, and immunology, but is also incredibly relevant to promoting human health and treating a wide range of diseases, from infection to autoimmunity to metabolic disorders.