Immunology: You Remind Me of a Microbe I Know

Abstract

The majority of bacteria found within the gut are commensals, although it is unclear whether these organisms can elicit systemic immunity. New research indicates that gut-microbiota-specific serum antibodies targeting an epitope conserved among Gram-negative bacteria can protect the host from systemic pathogenic infection.

The gut microbiota provide a variety of benefits to mammalian hosts, including protection from invading organisms,degradation of indigestible plant polysaccharides, and promotion of immune cell development [1–4]. While the majority of bacteria in the gut are non-pathogenic, they share common antigens with pathogenic bacteria that have the potential to induce immune responses. It has been appreciated that microbiota-reactive antibodies can be found in healthy humans without chronic inflammatory conditions [5,6] as well as in animals [7]. Antibodies play a crucial role in protecting the host from infection: by binding pathogen epitopes, antibodies can promote phagocytosis by innate immune cells, block the effects of toxins, and prevent infection by bacteria. The presence of gut-microbiota-reactive antibodies in the periphery of healthy individuals suggests that the host is primed to respond to organisms that may escape the gastrointestinal tract. However, given the overlap in antigens expressed by the microbiota and by pathogens, it remains unclear whether these microbiota-specific antibodies can protect against pathogen infection.

Antibodies come in many forms, referred to as isotypes. The most abundant antibody isotypes are immunoglobulin (Ig) M, A and G [8]. IgM is one of the first antibodies generated in response to a given antigen and is often low in affinity and specificity compared with the other isotypes. Both IgA and IgG are made later during an immune reaction: antibody-producing cells (B cells) require help from T cells to generate these highly specific antibodies. Importantly, while IgA antibodies are primarily found at mucosal sites, such as the gut, [9] IgG is most abundant within the serum. New research from the Nunez laboratory published recently in Immunity reports that serum IgG antibodies generated to target specific members of the microbiota can actively protect the host from systemic infection with pathogenic bacteria expressing the same epitope.

Zeng et al. [9] demonstrate the presence of circulating antibodies that react against fecal antigens in specific pathogen-free (SPF) mice. Germ-free (GF) mice offer a tool to test for microbiota-specific effects because these animals are born and reared in a completely sterile facility and are not colonized with any bacteria, viruses or fungi. Circulating antibodies that react against fecal antigens are not observed in GF mice; thus, the authors conclude that these serum antibodies are reactive to the microbiota.

While the presence of microbiota-specific antibodies has been previously described [5,7,10–13], the specific bacterial members of the microbiota that promote this antibody production are unknown. Utilizing deep sequencing of systemic sites such as the spleen and mesenteric lymph nodes, Zeng et al. [9] find that bacterial DNA from these sites was enriched from Gram-negative bacteria compared with the total bacterial populations found in the gut. Concurrent with this observation, microbiota-reactive IgG taken from serum was found to specifically bind epitopes from multiple Gram-negative but not Gram-positive bacteria, suggesting a protective role generated by and targeted towards Gram-negative bacteria within the gastrointestinal tract. Thus, a potential role for these antibodies might be protection of the host from organisms that translocate from the gut to the periphery. To test this hypothesis, SPF mice were treated with dextran sodium sulfate (DSS), a treatment that induces colitis-like symptoms.

In these experiments, mice are given DSS in their drinking water, which results in disruption of the epithelial barrier and penetration of bacteria to the systemic compartment. While no differences in microbiota-reactive serum IgG antibodies were identified, enhanced IgG concentrations and IgG binding of bacteria were detected in the feces. Critically, this antibody response was shown to protect hosts because B-cell-deficient mice (J_H^−/− mice) suffered higher loads of bacteria in their blood and succumbed to sepsis. These data indicate an important role for pre-existing microbiota-reactive antibodies in preventing the translocation of gut microbes.

In order to identify the bacteria that were being controlled by microbiota-reactive serum IgG antibodies, bacteria invading the spleens of J_H^−/− mice were analyzed. Indeed, Gram-negative bacteria in particular were enriched at these sites, indicating that, in the absence of antibody, Gram-negative organisms can easily breach the intestinal wall. In a key experiment, the authors sought to rescue increased bacteremia in J_H^−/− mice by infusing them with serum IgG isolated from wild-type SPF mice prior to infection with an Escherichia coli strain isolated from a J_H^−/− mouse spleen. Strikingly, SPF IgG was able to protect J_H^−/− mice from systemic E. coli challenge, definitively demonstrating that microbiota-reactive antibodies can control translocation and bacteremia of gut microbes.

What signals from these bacteria could be promoting the generation of antibody? Previous studies have shown the importance of Toll-like receptor (TLRs) in communication between microbiota and the immune system, including the development of microbiota-reactive antibodies [14–17]. TLRs are conserved receptors that recognize common molecules shared between microbial organisms, such as lipopolysaccharide and peptidoglycan [18]. Upon binding of these ligands by the appropriate TLRs, a cascade of inflammatory responses, including antibody production, is mounted against bacteria. To address the importance of TLR signaling in the production of systemic microbiota-reactive IgG antibodies, serum samples from TLR2^−/−, TLR4^−/−, and double-deficient TLR2^−/−TLR4^−/− mice were tested for reactivity against antigens derived from fecal microbiota. Mice deficient in TLR2 and TLR4 showed a marked reduction in production of antibodies against the microbiota. Similar to B-cell-deficient mice, the TLR2^−/−TLR4^−/− knockout mice were highly susceptible to DSS treatment compared with SPF mice and had a significant reduction in the IgG response to the microbiota. Zeng et al. [9] further show that TLR2^−/−TLR4^−/− signaling in B cells is required for the generation of microbiota-specific IgG by reconstituting J_H^−/− mice with wild-type or TLR2^−/− TLR4^−/− bone marrow. As J_H^−/− mice have no B cells, bone marrow transplantation from a donor restores a functional B-cell population within these knockout mice. The transplantation of wild-type bone marrow, but not TLR2^-−/−TLR4^−/−bone marrow, allows for the generation of microbiota-reactive IgG and protects animals against infection.

While there appear to be fairly high levels of microbiota-specific antibodies in SPF mice, it was not clear which antigen from the commensal population was generating this response. In a series of elegant experiments, Zeng et al. [9] identify one of the major antigens to elicit the production of microbiota-specific IgG. Isolating IgG from serum of SPF mice and using these antibodies to probe E. coli cytosol and outer membrane lysates on a western blot identified a major IgG-reactive antigen within the 7 kDa range from outer membrane preparations of E. coli. Previous studies have identified that a highly conserved outer membrane protein of this size is murein lipoprotein (MLP) [19]. A significant reduction in binding to bacterial lysate was observed when this IgG was tested for reactivity against E. coli lacking MLP. Additionally, this serum IgG was tested for reactivity against several Gram-negative bacteria, including pathogens like Salmonella and Klebsiella. MLP is a conserved lipoprotein expressed on several Gram-negative bacteria; it has been shown to play a role in the pathogenicity of Salmonella strains that can stimulate inflammatory responses [20]. The authors also find MLP-specific antibody in the serum of healthy human donors. In addition, the authors observed reduced binding of antibodies from human serum to E. coli lacking MLP, suggesting that MLP is the dominant antibody epitope. Injection of SPF mice with either E. coli expressing MLP or MLP-deficient E. coli demonstrated the importance of MLP as the recognized antigen: E. coli lacking MLP were found at higher numbers in the spleen compared with E. coli containing MLP. These differences were abolished in J_H^−/− mice, as both strains of E. coli were able to reach high levels in the spleen. However, J_H^−/− mice given MLP-specific IgG antibody were protected from E. coli infection.

This led to the key question of the study: can MLP-specific antibodies originally generated against the microbiota provide protection from pathogenic organisms expressing MLP? To this end, GF mice were immunized with a microbiota strain of E. coli that expressed or did not express MLP. Two weeks later, after developing antibodies against MLP, these GF mice were infected systemically with the Gram-negative pathogen Salmonella. While GF mice are normally highly susceptible to Salmonella infection, GF mice immunized with E. coli containing MLP were able to significantly reduce the amount of Salmonella in the spleen and liver. In contrast, GF mice given E. coli that did not express MLP had significantly higher levels of Salmonella in their organs. To validate the efficacy of the MLP-specific antibody in protecting against MLP⁺ pathogenic organisms, SPF mice given MLP-specific antibody prior to Salmonella infection were able to significantly reduce the loads of bacteria in their organs, confirming that the antibody provided protection from systemic infection (Figure 1).

Figure 1. Circulating microbiota-reactive antibodies against murein lipoprotein (MLP) can protect against pathogenic infection.

(Left) When the host is colonized with microbiota expressing MLP, antibodies are developed that can target bacteria expressing surface MLP. In the periphery, TLR4 signaling on B cells promotes the generation of MLP-specific antibodies. During bacterial infection with an MLP⁺ pathogen (here, Salmonella), MLP-specific antibodies are able to efficiently target the bacteria and prevent dissemination to peripheral tissues. (Right) In the absence of MLP-expressing microbiota or in the absence of TLR signaling, there is no production of MLP-specific antibodies. The absence of these antibodies leaves the host susceptible to systemic infection by MLP⁺ members of the microbiota and pathogenic bacteria.

While much work has been directed towards understanding how the host differentiates commensal bacteria from harmful pathogens, less work has been carried out to investigate how similarities between commensal and pathogenic bacteria can aid in protection from infection. Indeed, this work adds another layer to the beneficial effects of the microbiota. Colonization resistance can be seen as a first defense against invading pathogens and the generation of antibodies towards conserved antigens may be a secondary defense if pathogens or symbionts are able to escape the gastrointestinal tract and invade systemically. The new work by Zeng et al. [9] suggests that bacteria from the microbiota could act as a kind of ‘vaccination’ for other pathogenic organisms. The generation of MLP-specific antibodies in humans suggests that this mechanism may be at work already. Indeed, as the field continues to establish the concept of a ‘normal’ microbiota, the presence of specific organisms that can promote the generation of MLP-specific antibodies or antibodies against other conserved antigens may move us forward in our understanding of what a healthy microbiota looks like.